Right ventricular size and function under riociguat in pulmonary arterial hypertension and chronic thromboembolic pulmonary hypertension (the RIVER study)

This clinical investigation was performed as a retrospective analysis using the data of prospective, randomized, double-blind, multicentre, parallel-group, placebo-controlled clinical studies (riociguat trails PATENT-1 study, PATENTplus study, CHEST-1 study, phase II trial and Early Access Study) and the corresponding long-term extension trials with add-on of echocardiographic data obtained during routine assessments which have been performed during these trials. All riociguat studies were approved by the ethics committees of the respective centers. The ethics committee of the medical faculty, University of Heidelberg had no objection against the retrospective analysis of echocardiographic data within this study (S668–2015). Due to the small number of patients with available echocardiographic data receiving placebo, the data was evaluated as single armed, observational study.

German centres were contacted to include patients who were enrolled within the riociguat phase II and III trials and received routine echocardiographic assessments throughout the study period. In five centres Dresden, Hannover, Heidelberg, Giessen and Regensburg echocardiographic assessments were routinely performed at baseline and after 3, 6 and 12 months beside the parameters which have been obtained in the context of the riociguat trials. All patients with PAH (PATENT-1 and 2, PATENTplus) or CTEPH (CHEST-1 and 2, Early Access Study) randomized into one of the trials (i.e. for whom study medication had been assigned and a package had been opened) and who had at least once been administered study medication were eligible for the study, if echocardiography had been routinely performed at baseline and at least once during the course of the trial.

For analysis of the echocardiographic parameters including RA and RV areas, TAPSE, RV thickness, tricuspid regurgitation velocity, RV fractional area change, right and left ventricular systolic function, left ventricular eccentricity index, pulmonary artery diameter and presence of pericardial effusion we obtained the DICOM-films of the on-line echocardiographic assessments performed in each centre and analysed them off-line in Heidelberg. In case of low image quality and lacking multiple measurements, values were not included in the data base. For the off-line readings specialised software was used (TOMTEC) by an experienced physician. The investigator (AM) who performed the off-line readings of the echocardiography-DICOM-data was blinded to the patient name, treatment group, location and time point of the assessment, demographic and clinical data. All echocardiographic measurements were performed according to American Society of Echocardiography/European Association of Cardiovascular imaging guidelines [11]. Accordingly, RA area and RV area were measured at the end of ventricular diastole (largest volume), tracing following the endocardium from the lateral aspect of the tricuspid annulus to the septal aspect, excluding the area of the annulus and the leaflets (with regard to RA area) and excluding the area of the annulus and trabecular structures (with regard to RV area). RV systolic area was calculated with the same methodology but at the end of the systole. RV fractional area change was defined as: [(RV end-diastolic area – RV end-systolic area)/ RV diastolic area]*100. TAPSE was calculated using M-mode from the lateral tricuspid annulus. Results of off-line readings were forwarded to an external contract research organisation in order to merge with the clinical data of the riociguat trials. A clinically significant improvement of right heart size was defined as a ≥ 15% reduction RA or RV area [12].

Data from the riociguat studies were used for this study to complete clinical characterization of the patient cohort. The respective parameters from the riociguat trials database were merged with the echocardiography data after double pseudonymization for data protection by contract research organisation. Parameters comprised of baseline demographics, vital signs, WHO functional class, hemodynamics, 6-min walking distance, NTpro-BNP and PH targeted treatment.

Patients who received riociguat in the core study were included with their baseline examination of the core study and the subsequent follow-up examinations. Patients who received placebo in the core study and riociguat in the extension study were included with their final examination of the double-blind study (= baseline examination of the extension study) and the subsequent follow-up examinations.

Patients with invalid echocardiographic recordings, insufficient quality of recordings or missing baseline echocardiography were excluded from the analysis.

Descriptive statistics are displayed by mean ± standard deviation, median and interquartile ranges. Frequency tables are provided for qualitative data. The primary endpoint was defined as the change of RA area from baseline to 12 months of treatment. The second part of the primary endpoint with comparison of changes during treatment between riociguat and placebo group was not possible due to the small sample size of the placebo group (n = 5). Previous power analysis for the primary endpoint determined a statistical power of at least 87.7% at an alpha level of 0.025 with a sample size of 75 patients if the true treatment effect was a reduction of RA area of at least 3 cm² with a standard deviation of the difference of 7.5 cm² after 12 months of treatment.

For the comparison of baseline and follow up values the t-test for paired measurements or the Wilcoxon signed rank test was used, as appropriate. Frequency distributions between baseline and follow-up were compared with the test of marginal homogeneity if at least one of the variables had more than two different values. For 2 × 2 tables the McNemar test was performed. Comparisons between groups were analysed with the t-test for independent samples or the Mann Whitney U test. For the comparison of frequency distributions between groups Fishers exact test was used.

For missing values of echocardiographic parameters after 12 months, baseline observation carried forward (BOCF) and last observation carried forward (LOCF) methods were performed as sensitivity analysis.

Imaging quality was very good in > 90% of cases. In 71 patients with baseline-echocardiography, changes of echocardiographic parameters after three (n = 30), six (n = 34) and 12 months (n = 50) are shown in Table 2. Patients receiving riociguat therapy showed a significant reduction in RA area after 6 (− 2.4 ± 3.7 cm², 95% confidence interval (CI) -3.73, − 1.13; p < 0.001, n = 34) and 12 months (− 2.6 ± 4.4 cm², 95% CI -3.84, − 1.33; p < 0.001, n = 49). Likewise, patients showed a significant decrease in RV area (3 months: − 3.0 ± 3.0 cm², 95% CI -4.1, − 1.9; 6 months:− 3.0 ± 3.15 cm², 95% CI -4.23, − 1.96; 12 months: − 3.5 ± 5.2 cm², 95% CI -5.1, − 1.9; p < 0.001 for all time points; n = 29, 32, 44, respectively). Patients who completed all four echocardiographic assessments (at baseline, after 3, 6 and 12 months (n = 21)) presented with a significant reduction of RA and RV areas for all time points (Table 2, Fig. 2).

After three months seven of 30 patients had a marked reduction in RA area of ≥15% compared to baseline (23.3% of 30); after six months 14 patients reached this reduction (41.2% of 34) and after 12 months 23 patients (46.9% of 49) (Table 2).

At baseline 20 of 70 patients showed a RA area lower than 18cm² (28.6% of 70, one patient with baseline echocardiography had a missing RA area value). After 3 months 15 had a RA area < 18 cm² (50.0% of 30), after 6 months 18 (52.9% of 34) and after 12 months 25 (51.0% of 49), respectively (Table 2).

After 12 months of riociguat treatment patients showed a significant decrease in RV thickness (− 0.76 ± 2.2 mm, 95% CI -1.55, 0.03), tricuspid regurgitation velocity (− 0.36 ± 0.83 cm/sec, 95% CI -0.6, − 0.12), pulmonary artery diameter (− 1.64 ± 3.17 mm, 95% CI -2.92, 0.36) and a significant increase in TAPSE (2.95 ± 4.78 mm, 95% CI 1.52, 4.39) and RV fractional area change (8.12 ± 8.87 mm, 95% CI 4.61, 11.62) (Table 2).

Patients also presented with an improvement of qualitatively assessed RV systolic function after 12 months of treatment (normal systolic function at baseline n = 6 vs. 16 at 12 months, mild impairment 15 vs. 16, moderate impairment 15 vs. 7, severe impairment 9 vs. 6; χ²-test p = 0.016). Left ventricular systolic function and presence of pericardial effusion remained stable throughout the study period. Both LOCF and BOCF showed similar significance and lower effect sizes (data not shown).

Echocardiography after 12 months riociguat treatment was performed for 50 out of the 71 riociguat patients with echocardiography at riociguat start (70.4% of 71), 21 patients had no echocardiography after 12 months (29.6% of 71).

Patients who did not complete their 12-month echocardiographic assessment did not significantly differ regarding gender, age, height, weight, severity of the disease, 6MWD and hemodynamics from patients who received an echocardiographic examination after 12 months (Table 3, all p > 0.05). Baseline values for 6MWD and cardiac index were lower, pulmonary vascular resistance and NT-proBNP higher in patients who did not complete their 12 month examination, though differences between the two groups did not reach statistical significance (Table 3). Median values for PVR, cardiac index and cardiac output were similar between groups.

Patients with 12-month assessment showed a statistically significant but clinically not relevant lower diastolic blood pressure and higher PAWP (Table 3).

In 20 patients who received placebo treatment, data from baseline echocardiography was available. Out of them, five patients presented with echocardiographic data after 3 months. Patients who received placebo treatment had a mild increase in NT-proBNP (22 ± 104 pmol/l) and 6MWD (18 ± 55 m).

Due to the small sample size of five patients who received an echocardiographic assessment after 3 months, the effect of riociguat was not tested for statistical significance against the control group. Data of patients receiving placebo treatment are presented descriptively (Table 4). RA area slightly decreased within 3 months (0.4 ± 2.9), whereas RV area increased by 1.9 ± 6.4 cm². During the three month period, an increase was seen in TAPSE (0.5 ± 3.2 mm) and RV thickness (1 ± 0.8 mm). Median values of differences between baseline and 3 months did not reveal any marked changes. Presence of pericardial effusion and RV function also remained constant within the 3 months study period.

These data confirm the results of a recent unblinded single-center study [10] which showed a significant improvement of right ventricular area already after 3 months, an improvement of RV systolic function after 6 months and right atrial area after 12 months treatment. In the current study, RA area was already significantly improved after 6 months of riociguat treatment.

As the current guidelines on treatment goals in pulmonary hypertension state, RA area seems to be one of the most robust echocardiographic determinants of outcome [13] and strongly correlates with RV function [9]. Enlarged RA assessed by transthoracic echocardiography was significantly associated with lower transplant-free survival [7, 14] whereas RA area < 18 cm² was associated with a good prognosis [9, 14]. In this study a reduction of RA area ≥ 15% was achieved by almost 50% of the patients. It has been shown previously that a therapy response to PH targeted medication of ≥15% in RA and RV area may predict long-term outcome in patients with PH [12]. Using the proposed cut-off for a normal RA area of 15cm² in female and 16 cm² in males [15], 31% of patients who presented with baseline values above these cut-offs showed values within the normal ranges after one year of riociguat treatment.

A reduction of right heart size might be an indicator for an improvement of systolic function and reduction in volume overload. In this study the reduction of right heart size was accompanied by a significant improvement of RV systolic function assessed by RV fractional area change, TAPSE and qualitative assessment of RV systolic function.

Preclinical studies showed that Riociguat has a positive effect on pulmonary pressures [16], and is able to reverse RV remodeling [17]. In our study, riociguat administration was associated with an improvement of both RV systolic function, as well as of pulmonary pressures (TRV). The investigation of the underlying predominant mechanisms of riociguat treatment was however out of the scope of our study.

A strength of this study is the central off-line reading of echocardiographic data by a blinded experienced physician which standardized the interpretation of data and increases its quality. In addition, patients included in this analysis were enrolled in prospective randomized controlled riociguat trials, leading to a thorough and homogenous patient selection (due to similar in- and exclusion criteria of the trials) and high quality of clinical data. As the echocardiographic data derived from routine echocardiographic assessments and were assessed according to the local standards of the participating centres, the study results represent data from clinical practice.

On the other hand, the rate of patients who did not receive echocardiographic assessments after 12 months may have biased the results. However, sensitivity analyses of echocardiographic parameters after 12 months with BOCF and LOCF led to similar significant results with lower effect sizes (data not shown).

Further study limitations are the retrospective nature of the study, limited number of study centres and a small number of patients, in particular those receiving placebo treatment. A higher number of patients receiving riociguat would also have been desirable in order to compare the effects on patients with PAH vs. CTEPH.