Changes in cMRI parameters following a switch to riociguat from phosphodiesterase type 5 inhibitors (PDE5i) in patients with pulmonary arterial hypertension: a REPLACE substudy

Abstract

Abstract Background The REPLACE study investigated the effect of switching to riociguat (RIO) in patients with pulmonary arterial hypertension receiving PDE5i but still at intermediate risk. The centrally adjudicated composite primary endpoint was clinical improvement in the absence of clinical worsening, where clinical improvement was defined as meeting at least two of the following criteria: 6-minute walk distance (6MWD), increase by ≥10% or ≥30 m from baseline (BL) to Wk 24; World Health Organization functional class (WHO FC) I or II at Wk 24; or N-terminal prohormone of brain natriuretic peptide reduction of ≥30% from BL to Wk 24. Twice as many patients switching to RIO (45/111, 41%) met the primary endpoint compared with those remaining on PDE5i (23/113, 20%); odds ratio (OR): 2.78 (95% confidence interval [CI] 1.53–5.06); p=0.0007. Purpose Assess changes in right and left ventricular (RV; LV) function using cardiac magnetic resonance imaging (cMRI) in a subgroup of patients participating in REPLACE. Methods REPLACE was a randomised, open-label, 24-week, Phase 4 study (NCT02891850). Patients in WHO FC III, with 6MWD 165–440 m, were randomised to switch to RIO 2.5 mg–max tid or remain on PDE5i. Background endothelin receptor antagonist therapy was permitted in both arms. cMRI was performed on a subset of patients from the full analysis set as an exploratory substudy. The following parameters were measured at BL and Wk 24: RV and LV end-diastolic and end-systolic volumes (RVEDV; RVESV; LVEDV; LVESV), RV stroke volume and stroke volume index (RVSV; RVSVI), LV stroke volume (LVSV), RV ejection fraction (RVEF), and pericardial effusion. Results Twenty-seven patients participated in the cMRI substudy. This comprised 11/111 (10%) patients in the RIO arm (mean [standard deviation {SD}] 40.0 [12.4] years), and 16/113 (14%) patients (mean 44.5 [17.6] years) in the PDE5i arm. Like the main population, the treatment response in the cMRI subpopulation favoured RIO versus PDE5i (OR: 6.11 [95% CI 0.90–41.60]). From BL to Wk 24, RVEDV and RVESV decreased in the RIO treatment arm but increased in the PDE5i treatment arm (Table 1). Similar, but less pronounced, changes were observed for the left ventricle (LVESV, LVEDV). RVSV and RVEF levels were close to normal at BL and did not increase in either arm at Wk 24 (Table 1). Pericardial effusion, which was present in 5 patients in each group at BL, decreased in 1 patient in the RIO arm and no patients in the PDE5i arm. Conclusions Decreases in RVEDV and RVESV suggest improvements in cardiac function in the RIO arm compared with the PDE5i arm. Values for RVEF and RVSVI were close to normal at BL and did not change at Wk 24. Improvements in cMRI parameters were in line with the clinical improvement observed in patients switching to RIO in the overall population. Funding Acknowledgement Type of funding sources: Other. Main funding source(s): The REPLACE study was co-funded by Bayer AG (Berlin, Germany) and Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc. (Kenilworth, NJ, USA)