From Wiki Journal Club
Jump to navigation Jump to search
Redfield MM, et al. "Effect of Phosphodiesterase-5 Inhibition on Exercise Capacity and Clinical Status in Heart Failure with Preserved Ejection Fraction". JAMA. 2013. 309(12):1268-1277.
PubMedFull textPDF

Clinical Question

In patients with symptomatic heart failure and preserved ejection fraction (HFpEF, LVEF ≥ 50%, NYHA II-IV), does the phosphodiesterase-5 (PDE5) inhibitor sildenafil improve exercise capacity or clinical status compared to placebo?

Bottom Line

In patients with symptomatic HFpEF, sildenafil does not improve exercise capacity (measured by peak VO2) or clinical status at 24 weeks.

Major Points

Despite a large number of randomized controlled studies investigating the effects of neurohormonal blockade, beta blockade, vasodilation, and modulation of other common targets of cardiac therapeutics (see CHARM-Preserved, TOPCAT, NEAT-HFpEF) there is no therapy clearly proven to improve mortality or clinical status in patients with heart failure and preserved ejection fraction (HFpEF). By inhibiting the enzyme phosphodiesterase-5 (PDE5), PDE5 inhibitors have been shown to reverse adverse cardiac remodeling, enhance vascular, neuroendocrine, and renal function, and reduce pulmonary artery pressure, thus potentially targeting several key derangements thought to underlie HFpEF. Furthermore, PDE5 inhibition has previously been shown in a small study to improve hemodynamics, left ventricular diastolic function, left ventricular hypertrophy, and lung function after 6-12 months in patients with HFpEF.[1] A larger randomized controlled trial was needed to establish whether PDE5 inhibition may improve exercise capacity and clinical status in patients with HFpEF.

The 2013 Phosphodiesterase-5 Inhibition to Improve Clinical Status and Exercise Capacity in Heart Failure with Preserved Ejection Fraction (RELAX) trial randomized 216 patients with symptomatic HFpEF (LVEF ≥ 50%, NYHA II-IV) to the PDE5 inhibitor sildenafil versus placebo and assessed for a primary outcome of change in exercise capacity (assessed using peak VO2). At 24 weeks, patients randomized to sildenafil had no improvement in exercise capacity/peak VO2 compared to placebo. Secondary outcomes including 6-minute walk distance, a composite of general clinical status (death, hospitalization for cardiac or renal causes, or increased HF symptomatology), LV mass by cardiac MRI, and diastolic dysfunction parameters by echocardiogram were all also unchanged with sildenafil. Serious adverse events were similar in both groups although there was a small increase in vascular adverse events (headache, flushing, or hypotension) with sildenafil.

In summary, the results of RELAX suggest that PDE5 inhibition with sildenafil provides no benefit in terms of exercise capacity or general clinical status in patients with HFpEF. On the grounds of these data, PDE5 inhibitors have been relegated to the long list of ineffective therapies for HFpEF.


As of October 2017, no guidelines have been published that reflect the results of this trial.


  • Prospective, multi-center, double-blind, randomized controlled trial
  • N=216
    • Sildenafil (n=113)
    • Placebo (n=103)
  • Setting: 26 sites in the US and Canada
  • Enrollment: October 13, 2008 to February 21, 2012
  • Duration follow-up: 24 weeks
  • Analysis: Intention-to-treat
  • Primary outcome: Change in peak oxygen uptake on CPET (peak VO2)


Inclusion Criteria

  • Age ≥ 18 years
  • Previous clinical diagnosis of heart failure with current NYHA class II-IV symptoms
  • One of the following within the preceding 12 months
    • Hospitalization for acute decompensated CHF
    • Acute treatment for CHF with IV loop diuretic or hemofiltration
    • Chronic treatment with a loop diuretic for control of CHF symptoms and chronic diastolic dysfunction as evidenced by atrial enlargement on echocardiography
    • Mean PCWP > 15mmHg or LVEDP > 18 mmHg
  • LVEF ≥ 50% on echocardiogram or left ventriculogram within 12 months prior to consent
  • Stable medical therapy for 30 days including no addition, removal, or change in dosage by > 100% in CCB, ACEi/ARB, or beta-blocker
  • Either of
    • Peak VO2 ≤ 60% normal value with respiratory exchange ratio ≥ 1.0
    • One of the following
      • NTproBNP ≥ 400 pg/mL
      • NTproBNP < 400 with mean PCWP > 20 mmHg at rest or > 25 mmHg with exercise

Exclusion Criteria

  • Inability to perform exercise testing on cycle ergometer or walking in a hallway
  • Life expectancy < 1 year
  • Current or anticipated future need for nitrate therapy
  • Valve disease (> mild aortic or mitral stenosis or > moderate mitral or aortic regurgitation)
  • Hypertrophy or infiltrative cardiomyopathy
  • Pericardial disease
  • Pulmonary arterial disease
  • ACS, PCI, or CABG within 60 days
  • Morbid obesity or significant lung disease leading to dyspnea
  • Significant uncontrolled HTN or resting tachycardia
  • Severe renal dysfunction, liver dysfunction, or anemia
  • Childbearing age without effective contraception
  • Significant ischemia on stress testing
  • Listed for cardiac transplantation

Baseline Characteristics

All patients

  • Demographics: age 69, female 48%, BMI 32.9
  • CHF: NYHA II 47%, NYHA III 53%, MLFHQ score 43, HFH in past year 37%, LVEF 60
  • Co-morbidities: HTN 85%, ischemic heart disease 39%, AF 51%, DM 43%, COPD 19%, anemia 35%
  • Medications: Beta blocker 76%, ACE/ARB 70%, loop diuretic 77%, aldosterone antagonist 11%, CCB 31%, statin 64%
  • Labs: Cr 1.2, NTproBNP 700
  • CPET: peak oxygen consumption 11.7 (41% predicted), RER 1.09, chronotropic incompetence 77%, 6-minute walk distance 308 (69% predicted)


  • Randomized 1:1 to oral sildenafil or matching placebo
    • Randomization stratified by site and presence of atrial fibrillation
  • Participants who met entry criteria underwent baseline studies including history/physical, CPET, 6-minute walk test, clinical questionnaires, cardiac MRI, and biomarker testing
    • These studies were repeated at 12 and 24 weeks
  • Study drug was administered at a dose of 20MG three times daily for 12 weeks then 60MG three times daily for 12 weeks
  • If adverse effects were encountered, study staff could recommend discontinuation or reduction to a previously tolerated dose


Comparisons are sildenafil vs. placebo

Primary Outcomes

Peak VO2 (24 weeks, change from baseline)
-0.20 (95% CI -0.70 to 1.00) vs. -0.20 (95% CI -1.70 to 1.11) [p=0.90]

Secondary Outcomes

Clinical rank score
95.8 vs. 94.2 [p=0.85]
Change in 6-minute walk distance
15.0 (95% CI -26.0 to 45.0) vs. 5.0 (95% CI -37.0 to 55.0) [p=0.92]
0 (0) vs. 3 (3) [p=0.25]
Hospitalization for cardiovascular or renal cause
13 (13) vs. 15 (13) [p=0.89]
Change in Minnesota Living with Heart Failure Questionnaire score
-8 (95% CI -21 to 5) vs. -8 (95% CI -19 to 0) [p=0.44]

Subgroup Analysis

  • There was no trend toward improvement in peak oxygen consumption in any subgroup including patients still taking study drug at week 24, patients with or without LVH on cMRI, lower or higher pulmonary artery systolic pressure, lower or higher NTproBNP, with or without AF, on or not on statin, beta blocker, or ACEi/ARB therapy.

Adverse Events

Adverse events
78 (76) vs. 90 (80) [p=0.49]
Serious adverse events
16 (16) vs. 25 (22) [p=0.22]


  • Therapeutic sildenafil levels were associated with minimal increases in plasma cyclic guanosine monophosphate (and no statistically significant difference in the mean change in levels pre- and post-study drug), suggesting that PDE5 inhibitor mechanism may be impaired in patients with HFpEF. Thus, it is possible that a higher (supratherapeutic) dose may have resulted in higher cGMP levels leading to greater therapeutic effect.
  • Nearly 80% of patients enrolled had evidence of chronotropic incompetence, which may not be representative of the general HFpEF population. Patients without chronotropic incompetence may derive more benefit from PDE5 inhibition.
  • Patients were chosen for their ability to perform exercise testing, which may have introduced bias into the enrolled sample.


  • NHLBI-sponsored trial
  • Sponsor had no role in the design or conduct of the study, collection, management, analysis, or interpretation of the data, or preparation, review, or approval of the manusscript.

Further Reading