Therapeutic benefits of phosphodiesterase-5 inhibition in chronic heart failure: A meta-analysis.

BACKGROUND: Phosphodiesterase-5 inhibitors (PDE5i) have been shown to be beneficial for patients with pulmonary arterial hypertension. However, several studies would have documented a useful effect of PDE5i even for pulmonary hypertension secondary to left-sided chronic heart failure (CHF).
METHODS: We performed a meta-analysis including randomized controlled trials (RCTs) which had compared PDE5i (mostly sildenafil) and placebo in CHF patients.
RESULTS: Fourteen studies enrolling a total of 928 patients were incorporated in the meta-analysis. In heart failure with reduced left ventricular ejection fraction (HFREF), PDE5i, compared to placebo, significantly improved the composite of death and hospitalization (OR = 0.28; 95% CI: 0.10-0.74). They also improved peak VO2 [difference in means (MD): 3.76; 95% CI: 3.27-4.25], six-minute walking distance test (MD: 22.7 m; 95% CI: 8.19-37.21), and pulmonary arterial systolic pressure (MD: -11.52 mmHg; 95% CI: -15.56 to -7.49). Conversely, in CHF with preserved left ventricular ejection fraction (HFpEF), PDE5i proved not to yield any significant improvement of the investigated outcomes.
CONCLUSIONS: In HFREF, PDE5i showed beneficial effects on the composite of death and hospitalization, as well as on exercise capacity and pulmonary hemodynamics. Conversely, in HFpEF, no significant clinical, spiroergometric, or hemodynamic improvement was achieved using PDE5i therapy.

Background

The cardinal symptom of heart failure, i.e., the dyspnea, is largely attributable to pulmonary hypertension (PH) and congestion in the pulmonary vasculature [1]. So it is crucial to emphasize the very important role that PH plays in causing the symptoms and the clinical picture of heart failure either right-sided or left-sided or biventricular. PH associated with left heart disease (PH-LHD) coincides with the Group 2 of the most recent international classification of the PH [2] (Tables and . The favorable effects of phosphodiesterase-5 inhibitors (PDE5i), in particular sildenafil, in the treatment of PH are mainly attributed to the action exerted on the pulmonary arteriolar – precapillary district (so-called “precapillary pulmonary selectivity” of PDE5i) [3, 4]. In other words, the benefit of PDE5i in treating heart failure may originate from their hemodynamic effect for the combined post- and precapillary PH (Cpc-PH), but not for the isolated postcapillary PH (Ipc-PH) [5].

Table I

Comprehensive clinical classification of pulmonary hypertension

1. Pulmonary arterial hypertension	3. Pulmonary hypertension due to lung diseases and/or hypoxia
1.1. Idiopathic	3.1. Chronic obstructive pulmonary disease
1.2. Heritable	3.2. Interstitial lung disease
1.2.1. BMPR2 mutation	3.3. Other pulmonary diseases with mixed restrictive and obstructive pattern
1.2.2. Other mutations
1.3. Drugs and toxins induced	3.4. Sleep-disordered breathing
1.4. Associated with:	3.5. Alveolar hypoventilation disorders
1.4.1. Connective tissue disease	3.6. Chronic exposure to high altitude
1.4.2. Human immunodeficiency virus (HIV) infection	3.7. Developmental lung diseases
1.4.3. Portal hypertension	4. Chronic thromboembolic pulmonary hypertension and other pulmonary artery obstructions
1.4.4. Congenital heart disease
1.4.5. Schistosomiasis	4.1. Chronic thromboembolic pulmonary hypertension
1′. Pulmonary veno-occlusive disease and/or pulmonary capillary hemangiomatosis	4.2. Other pulmonary artery obstructions
	4.2.1. Angiosarcoma
1″. Persistent pulmonary hypertension of the newborn	4.2.2. Other intravascular tumors
2. Pulmonary hypertension due to left heart disease	4.2.3. Arteritis
2.1. Left ventricular systolic dysfunction	4.2.4. Congenital pulmonary arterial stenosis
2.2. Left ventricular diastolic dysfunction	4.2.5. Parasites (hydatidosis)
2.3. Valvular disease	5. Pulmonary hypertension with unclear and/or multifactorial mechanisms
2.4. Congenital/acquired left heart inflow/outflow tract obstruction and congenital cardiomyopathies
2.5. Congenital/acquired pulmonary vein stenosis	5.1. Hematological disorders: chronic hemolytic anemia, myeloproliferative disorders, and splenectomy
	5.2. Systemic disorders: sarcoidosis, pulmonary histiocytosis, lymphangioleiomyomatosis, and neurofibromatosis
	5.3. Metabolic disorders: glycogen storage disease, Gaucher disease, and thyroid disorders
	5.4. Others: pulmonary tumor thrombotic microangiopathy, fibrosing mediastinitis, chronic renal failure (with/without dialysis), and segmental pulmonary hypertension

Modified from Galiè et al. [2]

Aims

In this study, to evaluate the effects exercised by sildenafil or other PDE5i on some functional, hemodynamic, or clinical endpoints, a number of meta-analyses were separately conducted in patients with chronic heart failure (CHF) with reduced left ventricular ejection fraction (HFREF) or preserved left ventricular ejection fraction (HFpEF), respectively.

Methods

Study selection

A systematic search using some related terms was conducted using the PubMed and Embase electronic archives. We limited our search to adults (>18 years old) and to randomized controlled trials (RCTs). The study was performed according to the guidelines and recommendations expressed in the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) statement [6]. Search terms firstly included “heart failure,” “sildenafil,” “vardenafil,” “tadalafil,” “avanafil,” “udenafil,” “phosphodiesterase 5 inhibitors,” “phosphodiesterase type 5 inhibitors,” “PDE5 inhibitors,” “cardiac dysfunction,” and “pulmonary hypertension,” variously combined by means of the Boolean operators “AND” and “OR”. Roots and variants of the search terms were also used. Studies had to be prospective RCTs. In each of the studies admitted to meta-analysis, a comparison had to be made between a group of CHF patients taking a PDE5i and a second group assigned a placebo. Studies were incorporated in the meta-analysis provided that they had sufficient information about the explored hemodynamic and/or spiroergometric and/or clinical outcomes.

Study endpoints

The included RCTs were assessed for the following outcomes: exercise capacity [peak VO2 and six-minute walking distance (6MWD) test], cardiac performance [left ventricular ejection fraction (LVEF, %)], diastolic function (E/e′ ratio), and pulmonary resistance [mean pulmonary arterial pressure (mPAP), mmHg], pulmonary arterial systolic pressure (PASP, mmHg), and pulmonary vascular resistance (PVR, dyn·s cm−5). Clinical outcomes were assessed as all-cause death and hospitalization and adverse events.

Data extraction

All authors participated in determining the eligibility of candidate trials. The search included publications up to June 2016 and no lower date limit was applied. Titles and abstracts of all identified citations were reviewed independently by two authors (RDV and CA). Any candidate study was selected for further screening of the full text. In the event of a possible disagreement during data extraction, the intervention of a third reviewer (AC) was scheduled to solve any conflicting interpretation. Notably, it was decided that the studies selected for the meta-analysis should have included patients aged over 18 years. In addition, animal experimental studies as well as case reports of PDE5i administration without a control group were eliminated from the meta-analysis. Similarly, all studies not written in English, duplicated studies, review articles, editorials, and expert opinions were excluded.

Quality assessment

The authors assessed the risk of bias for the recruited RCTs using the Cochrane Collaboration Risk of Bias Tool. The following risks of bias were evaluated: (1) random sequence generation; (2) allocation concealment; (3) blinding of participants and personnel; (4) blinding of outcome assessment; (5) incomplete outcome data; and (6) other bias.

Statistical analysis

In the case of dichotomous variables, e.g., the composite of “death and hospitalizations” or adverse events, the effect size was expressed as odds ratio with a 95% CI, using Mantel–Haenszel method as the weighting method. When the endpoint was a continuous variable, such as “change in mPAP” or “change in 6MWD,” the effect size was expressed as a difference in means (MD) with a 95% CI, using inverse variance as the weighting method. Due to the large variety of patients, the effect size was calculated using a random effects model, even in case no heterogeneity was found. Statistical heterogeneity across studies was tested using Cochran’s Q test and I2 statistic (coefficient of variability due to inter-study variability). Statistical analyses were performed using RevMan 5.3 software (available from the Cochrane Collaboration; http//www.cochrane.org) and Stata version 10 (Stata Corp LP, College Station, TX, USA).

Results

In our meta-analyses, 14 studies were incorporated on the whole (Fig. , Tables and . Among them, 13 were RCTs [7–16, 18–20] and 1 was a subgroup analysis [17]. Patients affected by HFREF included in our meta-analysis were 555. All of them were derived from the pooling of 9 RCTs plus the aforementioned subanalysis study (Tables and . Conversely, patients with HFpEF included in our meta-analysis were 373 on the whole. This value corresponds to the sum of the patients enrolled by 4 RCTs [8, 11, 14, 19], specifically aimed to explore the effects of PDE5i in HFpEF.

Fig. 1.

Flow diagram for meta-analysis according to PRISMA statement

Table III

Baseline features of included RCTs

		Amin et al. (2013) [7]	Andersen et al. (2013) [8]	Behling et al. (2008) [9]	Guazzi et al. (2011) [10]	Guazzi et al. (2011) [11]	Guazzi et al. (2012) [12]	Guazzi et al. (2007) [13]	Hoendermis et al. (2015)[14]	Katz et al. (2005) [15]	Kim et al. (2015) [16]	Lewis et al. (2008) [17]	Lewis et al. (2007) [18]	Redfield et al. (2013) [19]	Webster et al. (2004) [20]
Subjects randomized (n; PDE5i/placebo)		53/53	35/35	11/8	23/22	22/22	16/16	23/23	21/22	60/72	21/20	15/15	17/17	113/103	35/35
Drug name		Sildenafil	Sildenafil	Sildenafil	Sildenafil	Sildenafil	Sildenafil	Sildenafil	Sildenafil	Sildenafil	Udenafil	Sildenafil	Sildenafil	Sildenafil	Sildenafil
Drug dosage		25 mg bid for first 2 weeks	40 mg tid	50 mg tid	50 mg tid	50 mg tid	50 mg tid	50 mg bid	20 mg tid for first 2 weeks	25/50/100 mg	50 mg bid for first 4 weeks	25–75 mg tid	25–75 mg tid	20 mg tid for first 12 weeks	50 mg once daily
		50 mg tid for next 10 weeks							60 mg tid for next 10 weeks		100 mg bid for next 8 weeks			60 mg tid for next 12 weeks
Inclusion criteria		HFREF	Diastolic dysfunction after MI	HFREF	HFREF	HFpEF with PH	HFREF with PH and EOB	HFREF	HFpEF with PH	CHF (HFREF) with ED	HFREF	HFREF with PHa	HFREF with PH	HFpEF	CHF (HFREF)
Entry criteria	NHYA	II–III	–	I–III	II–III	II–IV	III–IV	II–III	II–III	I–III	II–IV	II–IV	II–IV	II–IV	II–III
	LVEF	<35%	≥45%	≤40%	<40%	≥50%	<45%	≤45%	≥45%	≤40%	≤40%	<40%	<40%	≥50%	–
Follow-up duration (months)		3	2	1	12	12	12	6	3	3	3	3	3	6	1.5
Outcome measures		BP, NYHA, and 6MWD	Echo-, cardiac catheterization, CPET, and 6MWD	Echo-, CPET, and FMD	Echo-, CPET, BNP, and QoL	Echo-, cardiac catheterization, and QoL	CPET and cardiac catheterization	Echo-, CPET, and FMD	Cardiac catheterization, CPET, and Echo-	International index of erectile function	Echo- and CPET	Cardiac catheterization, CPET, and ventriculography	CPET: peak VO2	Echo-, CMRI, CPET, and 6MWD	International index of erectile function

CHF: chronic heart failure; HFREF: heart failure with reduced left ventricular ejection fraction; HFpEF: heart failure with preserved left ventricular ejection fraction; PH: pulmonary hypertension; EOB: exercise oscillatory breathing; MI: myocardial infarction; NYHA: New York Heart Association; PDE5i: phosphodiesterase-5 inhibitor; CPET: cardiopulmonary exercise test; echo-: echocardiography; FMD: flow-mediated dilatation; BNP: B-type natriuretic peptide; QoL: quality of life; BP: blood pressure; 6MWD: six-minute walking distance test; CMRI: cardiac magnetic resonance imaging

Subanalysis of Lewis et al. (2007) [18]

Therefore, a total of 928 patients with CHF were considered for the elaboration of the meta-analyses conducted in the course of our research. Among the included studies, 444 patients were assigned to sildenafil (with 443 patients assigned to placebo), and 21 patients were assigned to udenafil (with 20 patients assigned to placebo) (Tables and .

Clinical outcomes (death and/or hospitalizations, adverse events)

Seven RCTs of HFREF [7, 12, 13, 15, 16, 18, 20] reported clinical outcomes, with 5 hospitalization events occurring in the PDE5i arm and 17 occurring in the control arm. These results indicate a significant benefit conferred by PDE5i against hospitalization (OR = 0.28; 95% CI: 0.10–0.74; p = 0.03; Fig. ). Among the three RCTs concerning HFpEF that had included the endpoints of death and hospitalizations, one study [11] did not report any event, whereas the remaining two studies [14, 19] signaled 16 hospitalization events on the whole occurring in the PDE5i arm and 18 occurring in the control arm (OR = 0.81; 95% CI: 0.41–1.63; p = 0.64; Fig. ). During the follow-up period, five deaths were reported.

Fig. 2.

Deaths/hospitalizations HF

The occurrence of adverse events in these studies did not significantly differ between the PDE5i arm and the control arm (Fig. ).

Fig. 3.

Adverse events in patients with CHF

Exercise capacity and cardiac performance

The use of PDE5i significantly improved exercise capacity in patients with HFREF (Figs and . In particular, among the six RCTs that had investigated the peak VO2 in HFREF patients [9, 10, 12, 13, 16, 18] this parameter was improved by the use of PDE5i (MD: 3.76; 95% CI: 3.27–4.25; p < 0.0001; Fig. ). Similarly, based on the results of two studies [7, 18], in HFREF patients PDE5i use yielded a significant improvement of 6MWD compared to placebo arm (MD: 22.7 m; 95% CI: 8.19–37.21; p = 0.002; Fig. ). By contrast, in the RCTs of patients with HFpEF no benefit ensued from PDE5i use regarding exercise capacity as measured by cardiopulmonary exercise test or 6MWD (Figs and .

Fig. 4.

Peak VO2 in CHF

Fig. 5.

Six-minute walking distance test in patients with CHF

As regards the assessment of LVEF in patients with HFREF, based on the results of four studies [10, 13, 16, 18], the use of PDE5i was associated with a significant increase in LVEF compared to placebo (MD: 4.30%; 95% CI: 2.18%–6.42%; p < 0.0001; Fig. ). By contrast, the use of PDE5i for HFpEF patients resulted only in a nonsignificant tendency for increased LVEF (MD: 2.28%; 95% CI: −0.35% to 4.91%; p = 0.10; Fig. ).

Fig. 6.

LVEF in HFREF and HFpEF patients under treatment with PDE5i

The use of PDE5i in HFREF decreased mitral annular E/e′ ratio, but did not significantly affect this parameter in HFpEF (Fig. ).

Fig. 7.

E/e′ ratio in HFREF and HFpEF patients

Pulmonary resistance and pulmonary pressures (Figs and )

For patients with HFREF, PDE5i caused a nonsignificant reduction in mPAP (MD: −6.73 mmHg; 95% CI: −14.37 to 0.91; p = 0.11), whereas PASP was significantly reduced (MD: −11.52 mmHg; 95% CI: −15.56 to −7.49; p < 0.001; Fig. ).

Fig. 8.

Pulmonary pressures in CHF patients

The PDE5i-mediated improvement in pulmonary hemodynamic parameters for patients with HFREF was concordant among the RCTs. The use of PDE5i proved not to be associated with any significant improvement in pulmonary hemodynamics in patients with HFpEF (Figs and ; however, the included RCTs showed very high heterogeneity (I2: 99% for both mPAP and PASP in HFpEF patients; Fig. ).

PVR during therapy with PDE5i

Discussion

The illustration of the various studies centered around the PDE5i use in heart failure is far from simple. In addition, to explain the substantial failure of PDE5i in HFpEF, you may need to refer to specific categories of hemodynamic profile regarding the pulmonary circulation. However, such an approach is only applicable to RCTs in which pulmonary catheterization was performed (5 out of 13; see Tables and ). Some aspects of this issue are highlighted below.

Table IV

Different impact of PDE5 inhibitors according to pulmonary hemodynamics

		Amin et al. (2013) [7]	Andersen et al. (2013) [8]	Behling et al. (2008) [9]	Guazzi et al. (2011) [10]	Guazzi et al. (2011) [11]	Guazzi et al. (2012) [12]	Guazzi et al. (2007) [13]	Hoendermis et al. (2015) [14]	Kim et al. (2015) [16]	Lewis et al. (2008) [17]	Lewis et al. (2007) [18]	Redfield et al. (2013) [19]
	Inclusion criteria	HFREF	Diastolic dysfunction after MI	HFREF	HFREF	HFpEF with PH	HFREF with PH and EOB	HFREF	HFpEF with PH	HFREF	HFREF with PHb	HFREF with PH	HFpEF
Pulmonary hemod ynamic parameters	mPAP (mmHg: PDE5i/placebo)	–	19/20	36.2/39.8a	24.6/25.2a	39/37	35/34	22.7/21.5a	35/35	27/28.2a	30/33	30/33	27/27a
	dPAP (mmHg; PDE5i/placebo)	–	14/15	–	–	31.6/29.7	–	–	20/21	–	–	–	–
	PCWP (mmHg; PDE5i/placebo)	–	12/13	–	–	22/21.9	21/20	–	19.9/20.8	–	18/19	18/19	–
	TPG (mmHg; PDE5i/placebo)	–	7/7	–	–	16.2/14.5	15.2/14.7	–	13/13	–	12/14	12/14	–
	DPG (mmHg; PDE5i/placebo)	–	2/2	–	–	9.6/7.8	–	–	2/−1	–	–	–	–
	PVR (dyn·s·cm−5; PDE5i/placebo)	–	207/220	–	–	310.4/261.6	360/354	–	207/203	–	340/360	340/360	–
Features of combined post- and precapillary PH (DPG ≥7 mmHg; PVR >3 WU [>240 dyn·s·cm−5])		Not investigated	No	Not investigated	Not investigated	Yes	Yes	No	Mainly no (Cpc-PH in 12%)	Not investigated	Yes	Yes	Not investigated
Outcomes	Exercise capacity	No change	No change	Improved	Improved	N/A	Improved	Improved	No change	Improved	Improved	Improved	No change
	LV function	N/A	Improved	No change	Improved	Improved	Improved	Improved	No change	Improved	N/A	Improved	No change
	Pulmonary pressure	N/A	No change	Reduced	Reduced	Reduced	Reduced	Reduced	No change	Reduced	Reduced	Reduced	No change

Improvement in exercise capacity was evaluated based on the changes in peak VO2 and VE/VCO2 slope evidenced by cardiopulmonary exercise test or based on 6MWD. Improvement in LV function was evaluated based on the changes in LVEF. Reduction in pulmonary pressures was evaluated based on the changes in mPAP, PCWP, and PVR by means of cardiac catheterization, or using PASP derived from echocardiogram

HFREF: heart failure with reduced left ventricular ejection fraction; HFpEF: heart failure with preserved left ventricular ejection fraction; PH: pulmonary hypertension; EOB: exercise oscillatory breathing; MI: myocardial infarction; mPAP: mean pulmonary arterial pressure; dPAP: diastolic pulmonary arterial pressure; PCWP: pulmonary capillary wedge pressure; TPG: transpulmonary gradient; DPG: diastolic pulmonary gradient; PVR: pulmonary vascular resistance; N/A: not applicable

Converted from echocardiographic PASP by the following equation: mPAP (mmHg) = [0.61 × PASP (mmHg)] + 2 mmHg [5]

Subanalysis of Lewis et al. (2007) [18]

Favorable effects of PDE5i in the subset of HFREF patients

First, the PDE5i have proven to improve the composite of death and hospitalizations compared to placebo in HFREF patients. This has to be emphasized because based on seven studies [7, 12, 13, 15, 16, 18, 20], it testifies the existence of an important protective role of PDE5i against the risk of death and hospitalizations in HFREF patients. Among the studies incorporated in the meta-analysis, sildenafil was used in six studies and udenafil in one, with a total of 460 patients investigated about the endpoint “death and hospitalizations” (see Fig. ). It should be noted that a significant effect on this “hard” endpoint was not achieved by any of the individual studies considered. (Notably, two studies were not evaluable for the absence of events, i.e., lack of death or hospitalization in both the arm of PDE5i-treated patients and the one of controls.) Therefore, a statistically significant protective effect against death and/or hospitalizations (odds ratio: 0.28; 95% CI: 0.10–0.74) was inferred in HFREF patients exclusively on the basis of the overall analysis of the aggregate data. However, this result has to be reported with the due emphasis because it is a novelty, and because it helps us to propose with the due caution the PDE5i, in particular sildenafil, as candidate drugs ready to be inserted into the group of drugs (ACE inhibitors, beta blockers, aldosterone receptor antagonists) that on the basis of substantial clinical evidence are currently regarded capable of providing significant benefit to patients with HFREF in terms of increased survival and/or survival free from hospitalizations. Obviously further studies, again in the form of RCTs, are warranted to corroborate and validate the results of this meta-analysis. As regards the functional parameters (exercise capacity and cardiac performance), a very important and solid evidence in favor of the use of PDE5i has emerged from our meta-analysis. Indeed a functional improvement, ensuing from the administration of PDE5i has been documented for the exercise capacity in HFREF patients. Indeed, based on six RCTs [9, 10, 12, 13, 16, 18] with a total of 206 HFREF patients randomized to PDE5i or placebo, a substantial improvement in the peak VO2 has been proven in the PDE5i-treated patients. In particular, three studies have evidenced a significant increase in peak VO2. Moreover, the analysis of aggregated data has confirmed the existence of a statistically significant meaning of the increase in peak VO2 in the entire study population, related to the use of PDE5i (weighted MD: 3.76; 95% CI: 3.27–4.25).

Among patients with HFREF, the 6MWD has been assessed only in two studies, whose overall evaluation by means of meta-analysis has evidenced an increase in functional capacity in the PDE5i arm (Fig. . Even the LVEF was improved compared to placebo in HFREF patients taking therapy with sildenafil (Fig. .

In studies evaluating the measurements of the mPAP (two studies), PASP (four studies), and PVR (two studies), a significant reduction was consistently detected across the studies for each of these indexes in HFREF patients treated with PDE5i compared to those taking placebo.

The functional, hemodynamic, and clinical response of HFpEF patients to the PDE5i pharmacological inhibition: disappointing overall results that deserve further research

Differently from the substantially favorable response of HFREF patients to PDE5i administration, we did not observe any significant and consistent benefits conferred by PDE5i treatment for patients with HFpEF. The reasons for this unsatisfactory response are at the moment unclear. In this regard, there are elements of significant perplexity in the fact that at least two studies [10, 16] would have documented an improvement in diastolic function index known as E/e′ ratio in patients with heart failure treated with sildenafil [10] or udenafil [16]. In addition, the molecular and biochemical pathways of sildenafil and related drugs, such as detected in experimental animals, appear to actually be compatible with the hypothesis of a favorable effect by PDE5i on hemodynamic parameters and clinical outcomes of patients with HFpEF [21]. Conversely, with regard to the relatively low efficacy of PDE5i on hemodynamic and spiroergometric parameters, as well as on clinical outcomes in patients with HFpEF, as evidenced by some studies included in our meta-analysis [14, 19], this might depend on a possible predominance of the cases of Ipc-PH in these studies. This has been documented with certainty in the study by Hoendermis et al. [14], in which a condition of Cpc-PH, regarded as a crucial element for the occurrence of a comprehensive and effective pharmacodynamic action of PDE5i [5, 16] in the PH-LHD, was present only in 12% of cases. The fact that the HfpEF patients investigated in these studies were to be ascribed predominantly to the Ipc-PH category might have played a crucial role in the generation of disappointing results.

Therefore, the thesis aimed to support a useful effect limited to the HFREF patients, due to an alleged lack of efficacy of the PDE5i in HFpEF patients should be regarded not adequately proven yet [22]. In fact, the highlighted difference about the effects reported in the two echographic phenotypes might depend on a lower frequency of Cpc-PH profile in HFpEF patients rather than on a real critical role of the type of left ventricular dysfunction (HFREF or HFpEF) in determining the clinical efficacy of the PDE5i.

Therefore, to verify the possible causes of the unsatisfactory results of PDE5i in HFpEF, further studies, conducted by recruiting HFpEF patients belonging to the Cpc-PH category, would be warranted.

Study limitations

The results of this meta-analysis should be considered with caution because it has grouped data from a limited casuistry. In particular, data concerning the composite endpoint “death and hospitalizations” should be derived from a larger population before affirming the existence of an undoubtable advantage in terms of reduced mortality and hospitalizations in HFREF patients treated with PDE5i.

Conclusions

The use of PDE5i in patients with HFREF showed beneficial effects on pulmonary hemodynamics and exercise capacity. In addition, as regards the composite endpoint death/hospitalization, there was a significantly protective effect of PDE5i, limited to the HFREF patients.

Notably, the results concerning the composite endpoint death/hospitalization would require to be considered with caution, awaiting subsequent randomized trials and possible further meta-analyses to provide additional convincing evidence of therapeutic benefit in terms of reduced mortality and hospital admissions for HFREF patients undergone therapy with PDE5i.

Conversely, the use of PDE5i in patients with HFpEF showed disappointing results.

In fact, in the case of HFpEF patients, no significant improvement was achieved for each of the investigated endpoints (either functional, hemodynamic, or clinical).

However, the hypothesis that the unfavorable results detected in HFpEF patients might have been caused by a not proper selection of the patient population (i.e., paucity of the cases of combined post- and precapillary PH in the studies conducted to date) should be taken into account. Thus, further studies with well-defined pulmonary hemodynamic profile, including an adequate number of HFpEF patients with Cpc-PH, would be warranted to better clarify the real therapeutic potential of PDE5i even for treatment of HFpEF patients.

Table II

Hemodynamic features of the different clinical pulmonary hypertension groups

Definition	Characteristics	Clinical group(s)
PH	mPAP ≥ 25 mmHg	All
Precapillary PH	mPAP ≥ 25 mmHg	(1) Pulmonary arterial hypertension
	PCWP ≤ 15 mmHg	(3) PH due to lung disease
		(4) Chronic thromboembolic PH
		(5) PH with unclear and/or multifactorial mechanisms
Post-capillary PH	mPAP ≥ 25 mmHg
Isolated post-capillary PH (Ipc-PH)	PCWP > 15 mmHg	(2) PH due to left heart disease
Combined post- and precapillary PH (Cpc-PH)	DPG < 7 mmHg and/or PVR ≤ 3 WU
	DPG ≥ 7 mmHg and/or PVR > 3 WU	(5) PH with unclear and/or multifactorial mechanisms

DPG: diastolic pressure gradient (diastolic PAP – mean PCWP); mPAP: mean pulmonary arterial pressure; PCWP: pulmonary arterial wedge pressure; PH: pulmonary hypertension; PVR: pulmonary vascular resistance; WU: Wood units