A step forward in resolving an old issue: treatment of heart failure with preserved ejection fraction and renal dysfunction?

Heart failure with preserved ejection fraction (HFpEF) presently accounts for nearly 50% of all heart failure (HF) patients, and its prevalence is increasing.1,2 In comparison to heart failure with reduced ejection fraction (HFrEF), patients with HFpEF have a similar or higher risk of cardiovascular or all-cause mortality and non-cardiovascular hospitalization,3 which correlates with a considerable burden of comorbidities.4,5 Chronic kidney disease (CKD), defined as an estimated glomerular filtration rate (eGFR) of less than 60 mL/min/1.73 m2, affects up to 60% of HFpEF patients.6,7 Several studies have suggested that the presence of CKD outlines a distinct HFpEF phenotype characterised by a high risk of adverse outcomes.8–10 The excessive activity of the renin–angiotensin–aldosterone system (RAAS) plays a key role in the pathogenesis of both HF and renal dysfunction. Although evidence suggests that RAAS inhibition could prevent adverse events in CKD,11,12 treatment with RAAS inhibitors had a neutral effect on survival outcome in randomised clinical trials (RCTs) of unselected HFpEF patients.13–16 However, there is a possibility that RAAS inhibition could be beneficial for HFpEF patients with CKD, by targeting mechanisms that underlie the development of both HF and renal dysfunction. Currently, there is a paucity of data on the efficacy of RAAS inhibitors for the prevention of adverse outcomes in patients with HFpEF and CKD. This important matter was explored in a study presented in this issue of the European Journal of Preventive Cardiology by Tsujimoto and Kajio.17 This was an observational, retrospective analysis of the Treatment of Preserved Cardiac Function Heart Failure with an Aldosterone Antagonist (TOPCAT) trial data,16 including patients with HFpEF and mild-to-moderate CKD (i.e. eGFR <60 mL/min/1.73 m2 or urine albumin-to-creatinine ratio ≥30 mg/g). Patients with a eGFR of 30 mL/min/1.73 m2 or less were excluded. The analysis included 239 patients receiving RAAS inhibitors (i.e. angiotensin-converting enzyme (ACE) inhibitors, angiotensin receptor type1 blockers (ARBs) or both in 68.6%, 28.9% and 2.5% of patients, respectively) and 1227 patients who did not receive RAAS inhibitors. The two groups were compared with respect to the primary composite outcome of all-cause death, non-fatal myocardial infarction or stroke, or hospitalisation for HF. All-cause mortality, adverse cardiovascular events (i.e. cardiovascular mortality, non-fatal myocardial infarction/stroke) and hospitalisation for HF were analysed separately as secondary outcomes. The association between RAAS inhibitors and study outcomes was first analysed using multivariable regression models. Due to a significant imbalance in baseline characteristics, the authors have defined a propensity score (based on potential confounders for receiving RAAS inhibitors) that was used to match 231 pairs of patients with and without RAAS inhibitors. Survival analysis was then repeated in the propensity score-matched patients, as a sensitivity analysis. The main result was that RAAS inhibition significantly reduced the risk of the composite outcome of all-cause mortality, adverse cardiovascular events and HF hospitalisation by 25% (adjusted hazard ratio (HR) 0.75, 95% confidence interval (CI) 0.69–0.94; P = 0.01). This observation was confirmed in the propensity score-matched analysis (HR 0.67, 95% CI 0.50–0.90; P = 0.008). Improvement in primary outcome was comparable between ACE inhibitors and ARBs (adjusted HR 0.76, 95% CI 0.59–0.97; and adjusted HR 0.73, 95% CI 0.56–0.96, respectively), and there was no interaction with diabetes. The survival benefit was primarily driven by a reduction in all-cause and cardiovascular mortality by nearly 40%, while there was no significant reduction in rates of HF hospitalisation. On that basis, the authors have concluded that RAAS inhibitors can significantly reduce the risk of cardiovascular events and all-cause death in HFpEF patients with mild-to-moderate CKD. The strength of the present study lies in the inclusion of a contemporary HFpEF cohort from the TOPCAT trial, in which criteria for HFpEF and CKD were well specified and study outcomes were centrally adjudicated. Although the study was observational, a reduction in bias of non-randomised treatment allocation was achieved by the propensity score matching. However, propensity score analysis can only control for measurable confounders and, although it increases internal validity, it also reduces the generalisability of the results.18 Despite propensity score matching, uncontrolled and residual confounders may still be present, and therefore lack of randomisation is a major limitation to the present study. In addition, a small sample size might have underpowered some of the secondary analysis. Furthermore, the adverse effects of RAAS inhibitors (i.e. worsening renal function, hypotension and hyperkalaemia), with potentially serious consequences, have not been assessed.19 Likewise, target doses of RAAS inhibitors have not been specified, and these issues need to be addressed in future trials. Available evidence indicates a bidirectional relationship between HFpEF and CKD. For example, increased left ventricular (LV) filling pressure and right ventricular (RV) dysfunction in HFpEF lead to an elevated central venous pressure.20,21 In addition, fixed LV stroke volume and chronotropic incompetence result in an inadequate increase in cardiac output.22 These haemodynamic abnormalities are responsible for renal hypoperfusion that causes an excessive neurohormonal activation (including RAAS overactivity), which in the long run can have deleterious effects on both cardiac and renal function. Renal dysfunction can also promote HFpEF by worsening systemic inflammation and endothelial dysfunction.23 According to a novel paradigm of HFpEF pathogenesis, systemic inflammation and oxidative stress driven by comorbidities, including CKD, are responsible for cardiac microcirculatory endothelial dysfunction that leads to myocardial hypertrophy and dyastolic dysfunction.24 Indeed, the presence of CKD in patients with HFpEF was associated with significantly higher LV mass, more severe diastolic dysfunction, worse LV longitudinal and left atrial strain and RV dysfunction.7 In addition to ventricular remodeling, renal dysfunction in HFpEF has been associated with subtle LV systolic dysfunction, which may contribute to worse outcomes in HFpEF patients with CKD.6 RAAS inhibition has been shown to ameliorate some of the underlying pathophysiology. ACE inhibitors and ARBs have a beneficial effect on LV hypertrophy and fibrosis and protect against maladaptive remodeling.25,26 They also improve arterial compliance and endothelial dysfunction, and these effects lead to improvements in LV afterload, coronary blood flow and diastolic function.27,28 Of note, HFpEF patients have a higher risk of excessive falls in blood pressure and cardiac output with vasodilators,29 which is also pertinent to RAAS inhibitors. This adverse effect could offset the advantages of neurohormonal inhibition, particularly among the elderly, which stresses the importance of optimal dosing of RAAS inhibitors. Data on the clinical utility of RAAS inhibition in HFpEF diverge between RCTs and observational studies such as the study by Tsujimoto and Kajio.17 Considering RCTs, candesartan in the Candesartan in Patients with Chronic Heart Failure and Preserved Left-ventricular Ejection Fraction (CHARM Preserved) study had a neutral effect on the composite primary end-point of cardiovascular death or HF hospitalisation (adjusted HR 0.86, 95% CI 0.74–1.0; P = 0.051).13 Perindopril in the Perindopril in Elderly People with Chronic Heart Failure (PEP-CHF) study also had a neutral effect on the composite outcome of all-cause mortality or HF hospitalisation (HR 0.69, 95% CI 0.47–1.01; P = 0.055).15 A similar finding was observed in a more contemporary Irbesartan in Patients with Heart Failure and Preserved Ejection Fraction (I-PRESERVE) study, in which the primary outcome of all-cause death or cardiovascular hospitalisation was not significantly reduced with irbesartan (HR 0.95, 95% CI 0.86–1.05; P = 0.35).14 Finally, spironolactone failed to reduce cardiovascular mortality, aborted cardiac arrest or HF hospitalisation in the TOPCAT trial (HR 0.89, 95% CI 0.77–1.04; P = 0.14).16 In addition, several small RCTs failed to show an improvement in exercise capacity with either ACE inhibitors or ARBs compared with placebo.30–32 Taken together, RCTs have failed to demonstrate a survival benefit with RAAS inhibitors convincingly in HFpEF, although there was a trend towards improvement. Some trials have shown a modest reduction in hospitalisations for worsening HF,13,15,16 but there was no signal of harm with RAAS inhibitors.33 Furthermore, standard HF therapy including RAAS inhibitors appears ineffective to improve functional capacity and quality of life significantly in HFpEF.34 Conversely, endurance exercise training in conjunction with medical therapy has been shown to provide a clinically relevant improvement in exercise capacity in HFpEF patients.34 In contrast to RCTs, several observational studies and large registries of real-world HFpEF patients have consistently demonstrated a significant decrease in mortality with RAAS inhibitors, which supports the findings of the present study.35–39 In line with the results of the present study, RAAS inhibitors had no significant impact on HF hospitalisation in observational studies. A recent meta-analysis has confirmed lower mortality with RAAS inhibitors in observational studies of HFpEF patients, and the results were consistent irrespective of the LV ejection fraction cut-off used to define HFpEF (i.e. LV ejection fraction >40% vs. >50%), follow-up duration, the type of RAAS inhibitor used (i.e. ACE inhibitors vs. ARBs), or the statistical approach used to control the bias of non-randomised treatment allocation (i.e. multivariable adjustment vs. propensity score matching).40 In addition, observational data demonstrate that long-term neurohormonal blockade with RAAS inhibitors and beta-blockers could protect from the development of new-onset symptomatic HFpEF in high-risk elderly hypertensive patients.41 Several explanations have been put forward to clarify the discrepant findings between RCTs and observational studies. It has been suggested that RAAS inhibitors might be less effective because of a lesser degree of neurohormonal activation in HFpEF relative to HFrEF.42 In that case, frequent co-treatment with other neurohormonal inhibitors (i.e. beta-blockers, spironolactone) or the concomitant use of ACE inhibitors and ARBs in RCTs could have offset any incremental benefit from RAAS blockade in HFpEF. Notably, only a minority of patients (2.5%) in the present study used both ACE inhibitors and ARBs, which may have influenced the positive outcome. However, patients in observational studies also received beta-blockers and spironolactone in a proportion similar to RCTs, and this has not weakened the effectiveness of RAAS inhibitors in the observational setting. Moreover, a recent secondary analysis of the TOPCAT data provides new information on the prognostic relevance of neurohormonal overactivity in HFpEF by demonstrating that high resting heart rate is independently associated with adverse outcomes in patients in sinus rhythm, including those receiving beta-blockers.43 The other possibility is that, due to strict enrolment criteria, patients in RCTs were not truly representative of real-world HFpEF patients, which were more likely to be included in observational studies and registries. Indeed, a comparison of characteristics between patients from RCTs and observational studies has revealed that RCTs have enrolled younger and healthier patients, with better functional class and a lesser burden of comorbidities.40 This could explain lower than expected event rates of adverse outcomes that could have led to underpowering of RCTs to detect a difference in outcomes with and without RAAS inhibitors. Along the same lines, echocardiographic findings from recent RCTs have revealed that LV hypertrophy and severe diastolic dysfunction were less common than anticipated in HFpEF patients.44,45 This may explain why RAAS inhibitors, which can ameliorate LV hypertrophy and diastolic dysfunction, have failed to achieve positive outcomes in RCTs. In contrast to RCTs, observational studies have enrolled HFpEF patients of an older age and with a higher prevalence of comorbidities that might respond favourably to RAAS blockade, such as coronary artery disease, hypertension and atrial fibrillation, as shown in a recent meta-analysis.40 Of note, the present study population was characterised by advanced age (i.e. mean age ∼74 years) and a significant burden of comorbidities and, in that respect, appears similar to the real-world data. Most likely, comorbidities in HFpEF are not ‘innocent bystanders’, but may act as significant contributors to a worse prognosis.4 This is also true for CKD and, therefore, renoprotective and cardioprotective effects of RAAS inhibitors could benefit patients with HFpEF and CKD. In particular, in the presence of CKD, HFpEF patients express more severe LV hypertrophy and diastolic dysfunction.6,7 These cardiac abnormalities could be important targets for RAAS blockade and, in addition to renoprotection, this mechanism may form the basis for improved clinical outcomes observed in the study by Tsujimoto and Kajio.17 However, as this was a clinical study, it could not indicate the putative mechanisms by which RAAS inhibition contributed to positive outcomes. The study by Tsujimoto and Kajio is important as it provides novel insight into a potential survival benefit with ACE inhibitors and ARBs in patients with HFpEF and mild-to-moderate CKD.17 Although RCTs have a priority in a hierarchical order of scientific evidence, a well-performed observational study, such as the one by Tsujimoto and Kajio, may complement the evidence from RCTs and provide directions for future research. The results of the present study are relevant for clinical practice and need to be confirmed in larger and well-designed trials. These trials are also needed to clarify potential safety concerns with RAAS inhibitors in HFpEF and CKD, and to expand our knowledge on the utility of this treatment in patients with HF and mid-range ejection fraction and with more advanced CKD. The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article. The author(s) received no financial support for the research, authorship, and/or publication of this article.