Prognosis and Treatment With Phosphodiesterase 5 Inhibitors in Combined Post‐ and Precapillary Pulmonary Hypertension: A Propensity Score‐Matched Analysis From the Hellenic Pulmonary Hypertension Registry

ABSTRACT

Combined post‐ and precapillary pulmonary hypertension (CpcPH) comprises the most severe form of postcapillary PH. A severe precapillary component (pulmonary vascular resistance [PVR] > 5 WU) is critical for therapeutic decisions. Current treatment guidelines focus on optimizing underlying cardiac disease, while there are conflicting data regarding the efficacy and safety of pulmonary arterial hypertension (PAH) drugs in selected patients. This study examines the impact of PVR > 5 WU on survival in heart failure with preserved ejection fraction (HFpEF) and CpcPH and evaluates the effect of treatment with phosphodiesterase 5 inhibitors (PDE5is) on clinical and hemodynamic parameters and on prognosis. The Hellenic Pulmonary Hypertension Registry (HOPE) enrolls patients from all PH groups in Greece. This study focuses on Group 2 CpcPH patients with HFpEF. Propensity score matching was performed to reduce the risk of bias in the treatment selection and potential confounders. Kaplan–Meier curve was used to estimate 5‐year survival, and the log‐rank test was used for the comparisons. A total of 98 patients were included, with a median follow‐up of 2.9 years. PVR > 5 WU and age were independently associated with worse survival ([HR 2.15, 95% CI 1.13–4.83, p = 0.04], [HR 1.07, 95% CI 1.03–1.13, p = 0.003], respectively). Propensity‐matched cohort analysis indicated that PDE5i treatment was associated with a significant reduction in PVR at follow‐up (from median [IQR] 4.89 [1.9] WU to 3.1 [2.0] WU, p = 0.04) and a trend towards improved survival. Severe precapillary component is associated with impaired prognosis in CpcPH. While PDE5i treatment shows promise in improving hemodynamic outcomes, its effect on long‐term survival requires further investigation.

1. Introduction

Postcapillary pulmonary hypertension (PH) is a common complication of left heart disease [1, 2], severely affecting prognosis [1]. It is classified as Group 2 PH and is hemodynamically defined as mean pulmonary artery pressure (mPAP) > 20 mmHg and pulmonary arterial wedge pressure (PAWP) > 15 mmHg, according to 2022 ESC/ERS guidelines [3]. Its most severe form [4, 5, 6, 7, 8, 9], combined post‐ and precapillary pulmonary hypertension (CpcPH), is hemodynamically defined as mPAP > 20 mmHg, PAWP > 15 mmHg, and pulmonary vascular resistance (PVR) > 2 WU [3]. Several studies demonstrated the significance of PVR as a major prognosticator [9, 10, 11, 12], and a severe precapillary component in CpcPH is indicated when PVR exceeds 5 WU. Distinguishing this severe precapillary phenotype in the CpcPH population may be crucial for therapeutic decisions.

Current treatment recommendations of Group 2 PH focus on the optimization of medical management of the underlying cardiac disease [13, 14]. A potential acute reversibility of the precapillary component in CpcPH is proposed in some cases [15], suggesting a potential response to pulmonary arterial hypertension (PAH) drugs in selected patients [16]. There are conflicting data regarding the use of phosphodiesterase 5 inhibitors (PDE5is) in PH associated with heart failure with preserved ejection fraction (HFpEF) [3, 17, 18, 19, 20, 21].

This study seeks to investigate the impact of a severe precapillary component and treatment with PDE5is on the 5‐year survival of patients with HFpEF and CpcPH. Moreover, the impact of treatment with PDE5is on clinical and hemodynamic parameters has been evaluated.

2. Methods

The Hellenic Pulmonary Hypertension Registry (HOPE) is an ongoing national registry, initiated in January 2015, which actively enrolls patients with PH [22]. The HOPE has received approval from the Institutional Review Board of each of the 10 participating PH expert centers in Greece, in compliance with the Declaration of Helsinki. Patients have provided written informed consent for their inclusion in the registry. Data collection is facilitated through a web‐based platform (PAH tool by Inovultus Lda, Portugal), as previously outlined [22]. The project is endorsed by the Hellenic Society for the Study of Pulmonary Hypertension (HSSPH).

Patients registered up to November 20, 2023, were included. Inclusion criteria comprised a hemodynamic diagnosis of CpcPH (mPAP > 20 mmHg and PAWP > 15 mmHg or PAWP < 15 mmHg with a positive fluid challenge, and PVR > 2WU) associated with HFpEF as defined by either the 2015 [23] or 2022 [3] ESC/ERS guidelines, depending on the date of right heart catheterization (RHC) and age ≥ 18 years. HFpEF has been defined according to the 2021 ESC Guidelines on Heart Failure, as the presence of signs and symptoms of heart failure, a left ventricular ejection fraction (LVEF) ≥ 50%, and objective evidence of structural heart disease consistent with LV diastolic dysfunction, including elevated natriuretic peptides [24]. Other potential causes of PH were ruled out through various diagnostic procedures following a standard diagnostic algorithm.

Both incident and prevalent patients were included in the registry at the time of treatment initiation, which began within 3 months of a CpcPH diagnosis. An incident patient was defined as a patient who was newly diagnosed with CpcPH following RHC at one of the PH expert centers and enrolled in the HOPE between January 2015 and November 2023. Prevalent patients had been hemodynamically diagnosed with CpcPH before 2015 and were retrospectively enrolled in the registry.

World Health Organization Functional Class (WHO FC), six‐minute walk distance (6‐MWD), N‐terminal pro‐brain natriuretic peptide (NT‐proBNP), echocardiographic, and hemodynamic parameters were described at baseline evaluation. Pharmacological treatment and specific PDE5i pharmacotherapy at baseline (initiated within 3 months after PH diagnosis) were also recorded. Clinical and hemodynamic follow‐up occurred 3–6 months after the baseline assessment. Continuation rates for PDE5is, sodium‐glucose cotransporter 2 inhibitors (SGLT2is), diuretics, and therapies for other cardiovascular comorbidities were assessed throughout the follow‐up period through the national electronic health record. Survival data were obtained from patients' national health record or their relatives in case of death not yet registered in the national health record. Five‐year survival was defined as the primary event of the survival analysis. Patients lost to follow‐up or those with incomplete follow‐up were censored at the time of their last available medical contact.

Continuous variables following a normal distribution were presented as mean and standard deviation (SD), whereas variables that were not distributed normally were presented as median and interquartile range (IQR). The normality of a distribution was assessed by comparing the mean and median values, graphical representation of the distribution of the variables, and by using the Kolmogorov–Smirnov test. Qualitative variables were summarized using absolute and relative frequencies (n/N [%]). Statistical comparisons of continuous variables that exhibited a normal distribution were performed using the Student's t‐test, whereas the Wilcoxon rank‐sum test was employed for variables that did not follow a normal distribution. Categorical variables were compared with the χ ² test or the Fisher exact test if cell counts were small (≤ 5). The Kruskal–Wallis test was used for a comparison of continuous variables between more than two independent samples. Kaplan–Meier analysis was conducted to estimate free‐transplantation 5‐year survival. Log‐rank test was used to identify a statistical difference in survival between various groups. A multivariate Cox proportional hazards model was fitted to identify independent predictors of death. The purposeful selection process was used to identify candidate variables for our model. Variables with a p‐value < 0.2 in the univariate analysis were included in our model. The ratio of events to explanatory variables was 10 to avoid overfitting of the model. The application of a Cox proportional hazards model yielded hazard ratios and 95% confidence intervals. The Akaike Information Criterion (AIC) was used to select the model with the better fit (Table S1). To assess the robustness of the findings, several sensitivity analyses were conducted. First, patients with extreme NT‐proBNP values (< 5th or > 95th percentile) were excluded to determine the influence of outliers. A subgroup analysis was performed in patients who did not receive SGLT2i to examine the consistency of the effect of PVR > 5 WU within this subgroup. Additionally, a multivariable model, including fewer clinically significant variables such as age, PAWP, SGLT2i use, and PVR > 5 WU, was fitted to mitigate concerns about overfitting.

The study cohort was divided into two treatment groups. The control group consisted of patients who received the usual HFpEF treatment as per the judgment of the attending physician, namely optimal treated cardiovascular comorbidities (i.e., arterial hypertension, diabetes mellitus, dyslipidemia, and atrial fibrillation), diuretics in case of congestion, and a SGLT2i according to the latest ESC guidelines [23]. PDE5i group consisted of patients in whom a PDE5i was used on top of usual care. Propensity score matching was performed to reduce the risk of bias in the treatment selection and potential confounders.

A propensity score was estimated by fitting a logistic‐regression model adjusted for age, sex, WHO FC (I–II vs. III–IV), hypertension, diabetes, dyslipidemia, atrial fibrillation, obesity, NT‐proBNP, 6‐MWD, and hemodynamic parameters, namely, mean right atrial pressure (mRAP), mPAP, PAWP, PVR, cardiac index (CI), stroke volume index (SVi), and mixed venous oxygen saturation (SvO₂). Nearest neighbor matching without replacement was utilized to establish one‐to‐one pair matching between the two groups. Covariate balances were assessed by comparing standardized mean differences (SMDs) before and after matching. A SMD < 0.1 was deemed indicative of a minimal imbalance between the two groups. This approach was chosen to ensure optimal covariate balance between groups while minimizing bias. Higher matching ratios were opted out to avoid including poorly matched controls, which could compromise the internal validity of the analysis. In the matched cohort, survival curves were estimated using the Kaplan–Meier method, and for the survival comparison between the treatment strategies, the log‐rank test was used. A p‐value < 0.05 was considered statistically significant in this study. Data were analyzed using RStudio version 2023.03.0 + 386.

3. Results

3.1. Baseline Characteristics

In total, 98 patients (62% incident, 60% female) with CpcPH were included with a median (IQR) age at first visit of 69 (17) years. Figure 1 depicts a flow chart of how the cohort was reached. Most of the patients were in WHO FC II (47%) and III (44%), whereas systemic arterial hypertension was the most prevalent comorbidity (63%), with lower observed rates of obesity (38%), atrial fibrillation (35%), diabetes mellitus (32%), and dyslipidemia (19%). A comprehensive overview of the baseline characteristics of the patients according to PVR severity is presented in Table 1. Overall, median (IQR) PAWP was 20 (10) mmHg. Figure S1 indicates the whole PAWP values spectrum. Patients with PVR > 5 WU had higher NT‐proBNP values and a worse hemodynamic profile, that is, higher mPAP and lower CI, SVi, and SvO2 compared to those with a PVR ≤ 5WU. There was no difference in age, WHO FC, distribution of cardiopulmonary comorbidities, and 6‐MWD. Notably, 1/3 of patients with PVR > 5 WU received a PDE5i compared to 1/10 of patients with lower PVR (p = 0.002), whereas no difference was reported in the rest of the baseline medications. No endothelin receptor antagonists (ERAs) or other PAH drugs were used. Treatment with PDE5is, SGLT2is, and diuretics was maintained throughout the follow‐up period in all patients. Patients receiving PDE5is and SGLT2is remained on the same dosage regimen throughout the follow‐up period, with no adjustments or escalations in dose.

Figure 1.

Study's flow chart. Abbreviations: HFpEF, heart failure with preserved ejection fraction; HFrEF, heart failure with reduced ejection fraction; IpcPH, isolated postcapillary pulmonary hypertension; PDE5i, phosphodiesterase 5 inhibitor; PH, pulmonary hypertension; PVR, pulmonary vascular resistance; WU, woods units.

Table 1.

Baseline characteristics of CpcPH patients according to PVR group.

Characteristic	Overall, N = 98	PVR ≤ 5 WU, N = 52	PVR > 5 WU, N = 46
Age (years)	69 (17)	67 (17)	73 (12)
Female sex	60/98 (60)	32/52 (62)	28/46 (61)
BMI (kg/m2)	26 (9)	26 (10)	26 (7)
Incident	61 (62)	35 (67)	26 (56)
NYHA class
I	5/98 (5)	3/52 (5.8)	2/46 (4)
II	46/98 (47)	23/52 (44)	23/46 (50)
III	43/98 (44)	24/52 (46)	19/46 (41)
IV	4/98 (4)	2/52 (3.8)	2/46 (4.3)
Arterial hypertension	61/98 (62)	31/52 (59)	30/46 (65)
Diabetes mellitus	31/98 (32)	16/52 (30)	15/46 (33)
Dyslipidemia	19/98 (19)	6/52 (11)	13/46 (28)
Atrial fibrillation	34/98 (35)	17/52 (32)	17/46 (37)
Obesity	37/98 (38)	23/52 (45)	14/46 (30)
NT‐proBNP (pg/mL)	866 (1810)	569 (829)	2095 (3048)
6‐MWD (m)	351 (156)	389 (150)	305 (84)
Echocardiography
E/A ratio	1.35 ± 1.3	1.6 ± 1.1	1.1 ± 0.8
E/e′ ratio	18 ± 10	15 ± 9	21 ± 11
LAVi (mL/m2)	46 (18)	46 (28.2)	51 (15)
TAPSE (mm)	17 ± 4	18 ± 4	18 ± 5
Hemodynamics
mRAP (mmHg)	12.0 (7.8)	12.0 (9.0, 16.0)	12.0 (9.0, 15.8)
mPAP (mmHg)	42 (13)	40 (12)	47 (11)
PVR (WU)	4.5 (3.1)	3.4 (1.2)	6.5 (2.5)
PAWP (mmHg)	20 (10)	20 (8)	21 (10)
CI (mL/min/m2)	2.3 (1.1)	2.7 (2.2)	2.0 (1)
SVi (mL/m2)	31 (15)	35 (14)	28 (11)
SvO2 (%)	65 (13)	68 (12)	61 (12)
Medications
ACE‐i/ARB	30/98 (30.6)	16/52 (31)	14/46 (30.4)
β‐Blocker	27/98 (27.5)	11/52 (21)	16/46 (35)
MRA	25/98 (25.5)	13/52 (25)	12/46 (26)
SGLT2i	29/98 (29.5)	18/52 (35)	11/46 (24)
Diuretics	72/98 (74)	40/52 (77)	32/46 (69.6)
CCB	5/98 (5)	3/52 (6)	2/46 (4)
Digoxin	3/98 (3)	1/52 (2)	2/46 (4)
PDE5i	20/98 (20.4)	5/52 (10)	15/46 (32.6)
Tadalafil	8/98 (8.2)	2/52 (3.8)	6/46 (13)
Sildenafil	12/98 (12.2)	3/52 (5.8)	9/46 (19.5)

3.2. Impact of Severe Precapillary Component on Survival

During a 5‐year follow‐up period, eight patients were lost to follow‐up, none of whom were receiving PDE5i therapy. A total of 20 deaths occurred, corresponding to a mortality rate of 0.07 deaths per patient‐year, resulting in an overall 5‐year survival rate of 72.7% (Figure 2).

Figure 2.

Survival analysis of CpcPH patients' overall cohort. Abbreviation: CpcPH, combined post‐ and precapillary pulmonary hypertension.

Table 2 presents the univariate and multivariate Cox regression analysis of plausible prognostic factors. PVR > 5 WU and age were independently associated with worse survival ([HR 2.15, 95% CI 1.13–4.83, p = 0.04], [HR 1.07, 95% CI 1.03–1.13, p = 0.003], respectively). Results from sensitivity analyses are presented in Table S2 and demonstrate consistency with the primary findings, further supporting the robustness of the models.

Table 2.

Univariate and multivariate Cox regression analysis.

	Univariate				Multivariable
Characteristic	N	HRa	95% CIa	p‐Value	HRa	95% CIa	p Value
Age	90	1.07	1.02, 1.13	0.003	1.07	1.03, 1.13	0.003
Female sex	90	0.84	0.37, 1.88	0.70
BMI	90	0.97	0.89, 1.05	0.40
NYHA	90
I		—	—
II		0.75	0.09, 6.03	0.80
III		1.04	0.13, 8.09	> 0.90
IV		1.70	0.15, 19.4	0.70
Hypertension	90	2.09	0.95, 4.29	0.070
Diabetes	90	1.00	0.39, 2.55	> 0.90
Obesity	90	0.77	0.30, 1.96	0.60
AF	90	2.11	0.97, 4.61	0.060
Dyslipidemia	90	2.29	0.98, 4.04	0.080
NT‐proBNP	90	1.00	1.00, 1.00	0.60
6‐MWD	90	1.00	1.00, 1.01	0.20
E/A ratio	90	1.02	0.95, 1.09	0.50
E/e′ ratio	90	1.01	0.97, 1.06	0.60
LAVI (mL/m²)	90	1.03	0.97, 1.08	0.40
TAPSE (mm)	90	0.95	0.85, 1.07	0.40
mRAP	90	0.96	0.89, 1.05	0.40
mPAP	90	1.01	0.98, 1.05	0.40
PAWP	90	0.95	0.90, 1.01	0.074
PVR > 5WU	90	2.20	1.11, 4.90	0.035	2.15	1.13, 4.83	0.04
CI	90	0.93	0.64, 1.35	0.70
SvO2	90	1.01	0.97, 1.05	0.60
SVi	90	0.99	0.97, 1.02	0.70
PDE5i	90	1.44	0.60, 3.46	0.40
SGLT2i	90	0.71	0.30, 1.72	0.50

^a CI = confidence interval, HR = hazard ratio.

Figure 3 presents the Kaplan–Meier survival analysis among CpcPH patients with PVR > 5 WU and ≤ 5 WU. Patients with a severe precapillary component had a worse 5‐year survival compared to patients with PVR ≤ 5 WU (61.9% vs. 81.7%, p = 0.02).

Figure 3.

Survival analysis of CpcPH patients in different PVR groups. Red dashed line: overall survival. Abbreviations: CpcPH, combined post‐ and precapillary pulmonary hypertension; PVR, pulmonary vascular resistance; WU, woods units.

3.3. Baseline Characteristics of the Matched Cohorts

The propensity‐matching algorithm resulted in a match for 32 of the 98 patients. Table 3 outlines the baseline characteristics of the matched cohorts based on the treatment strategy, revealing no significant differences between the two groups. Among the 16 patients in the PDE5i group, 6 (38%) were treated with tadalafil (40 mg daily), whereas 10 (62%) received sildenafil (60 mg daily). Notably, 10 patients (62%) in the PDE5i group experienced side effects, including headaches in 7 patients, facial flushing in 2 patients, and minor nosebleeds in 1 patient.

Table 3.

Baseline characteristics of matched cohorts of patients with CpcPH.

Characteristic	Overall, N = 32	Control, N = 16	PDE5i, N = 16
Age (years)	71 (19)	68 (14)	72 (21)
Sex female	15/32 (47)	7/16 (44)	8/16 (50)
BMI (kg/m2)	25.8 (7.0)	25.4 (6.3)	26.7 (6.1)
NYHA class
I	1/32 (3.1)	0/16 (0)	1/16 (6.3)
II	14/32 (44)	8/16 (50)	6/16 (38)
III	15/32 (47)	7/16 (44)	8/16 (50)
IV	2/32 (6.3)	1/16 (6.3)	1/16 (6.3)
Hypertension	16/24 (67)	6/8 (75)	10/16 (63)
Diabetes	10/24 (42)	4/8 (50)	6/16 (38)
Dyslipidemia	4/24 (17)	1/8 (13)	3/16 (19)
Atrial fibrillation	10/24 (42)	3/8 (38)	7/16 (44)
Obesity	6/19 (32)	3/11 (27)	3/8 (38)
NT‐proBNP (pg/mL)	1797 (1647)	726 (1647)	2009 (1511)
6‐MWD (m)	369 (124)	369 (126)	316 (118)
Echocardiography
E/A ratio	1.1 ± 0.4	1.1 ± 0.4	1.2 ± 0.5
E/e′ ratio	13 ± 4	13 ± 4	13 ± 5
LAVi (mL/m2)	48 (14)	47 (13)	49 (14)
TAPSE (mm)	17 ± 3	17 ± 3	17 ± 4
Hemodynamics
mRAP (mmHg)	11.5 (7.3)	12.5 (5.3)	10.5 (9.3)
mPAP (mmHg)	45 (10)	45 (11)	45 (9)
PVR (WU)	5.39 (3.37)	5.95 (4.64)	4.89 (1.90)
PAWP (mmHg)	20 (9)	19 (14)	21 (7)
CI (mL/min/m2)	2.29 (1.22)	2.29 (1.12)	2.24 (1.31)
SVi (mL/m2)	32 (10)	33 (14)	32 (9)
SvO2 (%)	62 (13)	61 (12)	63 (57)
Medications
ACE‐i/ARB	10/32 (31.3)	6/16 (37.5)	4/16 (25)
β‐Blocker	8/32 (25)	3/16 (18.8)	5/16 (31)
MRA	8/32 (25)	4/16 (25)	4/16 (25)
SGLT2i	11/32 (34)	7/16 (4)	4/16 (25)
Diuretics	25/32 (78)	13/16 (81)	12/16 (75)
CCB	2/32 (6)	1/16 (6)	1/16 (6)
Digoxin	3/32 (9)	1/16 (6)	2/16 (6)

3.4. Impact of PDE5is on Functional Status and Hemodynamics

Figure 4 depicts the change of the functional (6MWD and NT‐proBNP) and hemodynamic status (PVR) of the 20 matched CpcPH patients treated with PDE5is. A decrease in PVR was observed in patients treated with PDE5is (median [IQR] 4.89 [1.9] WU vs. 3.1 [2.0] WU, p = 0.04). No significant difference was reported in 6MWD and NT‐proBNP.

Figure 4.

Changes in functional and hemodynamic status of CpcPH patients treated with PDE5is. Abbreviations: 6‐MWD, six‐minute walking distance; NT‐proBNP, N‐terminal‐pro‐brain natriuretic peptide; PVR, pulmonary vascular resistance.

3.5. Impact of PDE5is on Survival

Within the 5‐year follow‐up period, 10 deaths (31.2%) occurred (0.1 deaths per patient‐year) in the matched cohort. Figure 5 presents the survival of the matched cohort of CpcPH patients treated with PDE5i on top of usual care compared to the controls. Five‐year survival rate was 77.2% in the PDE5i cohort versus 37.0% in the control group (p = 0.07).

Figure 5.

Survival analysis of the matched cohort of CpcPH patients depending on the treatment group. Abbreviations: CpcPH, combined post‐ and precapillary pulmonary hypertension; PDE5i, phosphodiesterase 5 inhibitor.

4. Discussion

Over a span of 5 years, patients with PVR > 5 WU presented a 20% higher mortality rate compared to those with PVR ≤ 5 WU. Furthermore, an improvement in hemodynamics was observed in patients receiving a PDE5i on top of usual care for HFpEF. Although there was a trend towards improved survival in these patients, further studies are needed to confirm this finding.

Since 2022 [3], the use of the novel hemodynamic definition of PH has led to an increase in the number of patients diagnosed with CpcPH and enabled risk stratification of this population based on PVR [25]. Our study highlights the critical role of hemodynamics in the prognosis of CpcPH patients, consistent with previous research that established a PVR cut‐off of 5 WU as an indicator of significant precapillary component involvement [26, 27].

To date, there is no clear evidence supporting the use of PAH drugs in Group 2 PH. Almost all studies with ERAs in this population were negative reporting a high incidence of side effects [28]. 2022 ESC/ERS guidelines (3) do not recommend the use of PDE5i in isolated postcapillary PH, and they provide no definite recommendation in CpcPH. The existing evidence comes from small, short‐term studies or case series [1, 17, 19, 29, 30, 31, 32], and a small single‐center randomized clinical trial (RCT) in CpcPH, which reported a significant improvement of hemodynamic and echocardiographic indices in those who were assigned in the PDE5i group versus the placebo treatment group [20]. In contrast, the PASSION study, the only multicenter RCT in patients with HFpEF and CpcPH, did not manage to show favorable results in terms of heart failure hospitalization or all‐cause death for the tadalafil treatment group versus placebo [21]. However, the study was prematurely terminated due to the disruption in study medication supply and thus should be interpreted with caution. Other studies have shown that PDE5i treatment significantly improves both functional status and echocardiographic parameters after 1 year of therapy [18]. This is further supported by a retrospective analysis of the COMPERA registry, which included patients with HFpEF and a mean PVR of 7.0 ± 3.4 WU. The study suggested a potential benefit of PH therapies, particularly PDE5is, in this patient population [19]. These findings align with our results, which demonstrated a modest reduction in PVR.

Our study suggests that PDE5i treatment on top of usual care may benefit CpcPH patients by reducing PVR, although it did not demonstrate a clear survival benefit over the usual care, which could be attributed to the small sample size or lack of effect. Another limitation of our study is its retrospective design and the presence of other confounding factors that may have not been accounted for, despite the use of propensity score matching. Furthermore, given the limited sample size, our analysis necessarily included both incident and prevalent patients, as well as those receiving PDE5is and those who were not. This heterogeneity may have introduced potential confounding effects on survival outcomes. Although this is recognized as a limitation, it was essential to include all available patients to ensure sufficient statistical power. Additionally, the lack of detailed data on treatment escalation, particularly regarding diuretics, represents a limitation of this study. Escalation of diuretic therapy could serve as a marker of disease progression and may influence outcomes. The absence of this information restricts our ability to adjust for this potential confounder. However, we ensured that all patients receiving PDE5is or SGLT2is were maintained on consistent dosages throughout the study period, thereby minimizing confounding related to variability in these therapies. A potential referral bias is also acknowledged, as our registry exclusively comprised HFpEF patients monitored in specialized PH centers. This bias may exclude patients with HFpEF and potential CpcPH who were not referred to our center for RHC, potentially limiting the generalizability of our findings. Propensity score matching analysis was employed to minimize treatment selection bias.

In conclusion, the presence of a severe precapillary component in CpcPH patients is associated with a poorer prognosis. The addition of PDE5i treatment to standard care has the potential to improve the hemodynamic profile of patients with CpcPH associated with HFpEF. Further large‐scale RCTs are necessary to explore the potential survival benefits of this therapeutic approach.

Author Contributions

The conceptualization of the study was led by G.E.P. and A.A., with G.E.P. also responsible for the methodology and software development. Validation was carried out by A.A., S.‐A.M., and S.R., while G.E.P. conducted the formal analysis. The investigation was a collaborative effort involving G.E.P., A.A., S.‐A.M., P.G., T.C.‐T., A.M., A.B., A.N., E.D., P.K., D.T., A.A., A.M., A.Z., and G.G. G.E.P. managed the resources and data curation, and also prepared the original draft. A.A., S.‐A.M., and G.G. contributed to the review and editing process, with G.G. providing overall supervision. All authors have read and approved the final version of the manuscript.

Ethics Statement

The HOPE Registry received ethical approval from the Institutional Review Board of every participating center (AHEPA University Hospital, Thessaloniki; Mediterraneo Hospital, Athens; Onassis Cardiac Surgery Center, Athens; Hippokration University General Hospital, Athens; University Hospital of Ioannina) according to the Declaration of Helsinki, and all patients provided written informed consent for their inclusion in the study (Protocol Number 24643/3.6.2019).

Guarantor

George Giannakoulas accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.

Conflicts of Interest

G.G. has received fees for lectures and/or consultations from Actelion/Janssen, Bayer, Boehringer‐Ingelheim, ELPEN Pharmaceuticals, Ferrer‐Galenica, GlaxoSmithKline, Gossamer‐Bio, MSD, Pfizer, Lilly, and United Therapeutics. A.N. has received minor support via the University of Ioannina for participation in RCTs, advisory boards, and lecture fees from Amgen, Astra, Bayer, BMS, Boehringer Ingelheim, Janssen, Merck, Novartis, Novo Nordisk, Pfizer, Sanofi, and Servier, and has also received Horizon2020 funding for the CardioCare project. E.D. and P.K. received honorarium and consultation fees from Actelion Pharmaceuticals‐Janssen Hellas, MSD Merck Hellas, Elpen, and Galenica. D.T. has received in the past travel awards from Actelion/Janssen. A.M. has served on advisory boards for AstraZeneca, Bayer, Elpen, Janssen, MSD, and Novartis. S.R. has received fees for lectures and/or consultations from Abbott, Acceleron, Actelion, Aerovate, AOP, AstraZeneca, Bayer, Boehringer‐Ingelheim, Edwards, Ferrer, Gossamer, Janssen, MSD, and United Therapeutics. His institution has received research grants from Actelion, AstraZeneca, Bayer, Janssen, and Lempo. G.E.P., A.A., S.‐A.M., P.G., T.C.‐T., A.M., A.B., A.A., and A.Z. declare no conflicts of interest.

Supporting information

Supplementary material.