Mineralocorticoid receptor antagonists in heart failure: they work better when patients use them.

Abstract

This article refers to ‘Factors associated with underuse of mineralocorticoid receptor antagonists in heart failure with reduced ejection fraction: an analysis of 11 215 patients from the Swedish Heart Failure Registry’ by G. Savarese et al., published in this issue on pages 1326–1334. Prior to the 1980s, treatment options for patients with heart failure (HF) were limited in both number and their ability to improve outcomes. Beginning with pioneering studies using beta-blockers and angiotensin-converting enzyme inhibitors, clinical trials carried out over the past four decades have demonstrated the ability of a growing number of drugs and devices to reduce hospitalizations, prolong survival and improve quality of life in HF patients.1 Despite the fact that treatment options are now vastly greater than they were just a few short decades ago, HF-related morbidity and mortality have been reduced less than might have been anticipated from the results of these trials.2, 3 Explanations for why the cumulative impact of currently available therapies on the clinical course of HF patients has been less than expected can be attributed to several factors. Inclusion in the trials was restricted to willing patients who were stable and who could fulfil stringent entry criteria.4, 5 While this strategy enhances the likelihood of a successful study, it tends to overestimate the impact a therapy will have in a broader, more inclusive HF population. In addition, the drugs and devices that improve outcomes have been shown to do so only in stable patients with HF with reduced ejection fraction (HFrEF). To date, evidence that treatment strategies improve the survival of either patients who are hospitalized with decompensated HF or the >50% of the HF population with preserved ejection fraction is lacking. Even in HFrEF patients, the magnitude of improvement in morbidity and mortality has been far less than anticipated from clinical trial results. A major reason for this is that many effective HF therapies either are not being given to eligible patients or (in the case of drugs) are prescribed at suboptimal doses.6, 7 The situation with mineralocorticoid receptor antagonists (MRAs) is an example of this latter point. In a series of well-designed clinical trials carried out over the past three decades, the addition of an MRA to standard therapy significantly reduced hospitalizations and prolonged survival across a broad spectrum of symptomatic HFrEF patients.8-10 Moreover, the beneficial effects of MRAs were additive to those of other neurohormonal blocking agents and the drugs studied were shown to have an acceptable safety profile in the patients included in the trials. Based on these results, MRAs received a class I recommendation for the treatment of symptomatic HFrEF patients in the major HF guidelines.1, 11 However, despite strong guideline recommendations backed by compelling clinical trial evidence, MRA use has lagged substantially behind that of other HF drugs.12, 13 As a result, many HF patients who are at risk for hospitalization and death are not receiving a therapy that could help prevent events from occurring. In order to reverse this situation, it is essential that we understand both the magnitude of the problem and its causes. In this issue of the Journal, Savarese and colleagues provide further information about the extent to which MRAs are underutilized and they offer some insights into why this is the case.7 The investigators collected data on 11 215 patients with New York Heart Association (NYHA) class II–IV symptoms due to HFrEF [ejection fraction (EF) <40%] who were followed in the Swedish HF Registry between 2000 and 2012.7 Patients with documented contraindications to the use of MRAs were excluded so that the reported findings were based only on data obtained from patients who would be considered candidates for these drugs based on guideline recommendations. Moreover, recognizing that a variety of therapeutic strategies need to be initiated in HF patients, particularly in the period of time following the initial diagnosis, the investigators allowed 6 months from the time of entry into the registry before assessing MRA use. Despite careful attention to including only appropriate patients and allowing time for providers to initiate treatment, they found that only 40% of eligible HFrEF patients in the Swedish HF Registry were taking an MRA. Given the substantial and well recognized efficacy of these agents, the results are disturbing as they clearly document that an opportunity to reduce HF events was missed in this patient group. Before reaching broad conclusions about MRA utilization, however, it is important to consider the limitations of this database. A substantial number of patients in the Swedish HF Registry were excluded from this analysis based on missing data for MRA use, EF, NYHA class and HF duration. While the impact of these excluded patients on the results is uncertain, it is doubtful that it would have significantly biased the findings in one direction or the other. Perhaps more important is the fact that the MRA eligible HFrEF patients in the Swedish HF Registry averaged 75 ±11 years of age and it is possible that the advanced age of this population contributed to some extent to the low rate of MRA use. The Swedish HF Registry included a racially homogeneous group of patients living in a single country whose healthcare system differs (in some cases substantially) from that of other countries, so that these findings may not be generalizable to HFrEF populations in other regions. Regardless of these potential limitations, however, the consistency of the results with those derived from other populations reinforces the concept that there is a significant problem with the underutilization of MRAs in eligible HFrEF patients.12, 13 The 40% MRA utilization rate in the Swedish patients, however, is somewhat lower than in other study populations13 and there are some clues in the report to help explain why this might have occurred. The investigators found that lack of MRA use was associated with lack of need for diuretics, no cardiac resynchronization therapy/implantable cardioverter-defibrillator, higher blood pressure, no digoxin use and lower NYHA functional class, suggesting that these patients had milder HF. Although the class I recommendation for MRA use included NYHA class II–IV, conclusive evidence that patients with mild symptoms would benefit from an MRA was not available until the results of EMPHASIS-HF were published in 2011.10 However, when the investigators performed a consistency analysis, they found little change in MRA use in the period of time following publication of the EMPHASIS-HF results. In addition, the guideline recommendations for MRAs indentified patients with EF <35% while the Swedish HF Registry report included patients with EF <40%. Thus, providers may not have felt compelled to recommend MRA treatment to patients with milder symptoms (at least until late in the study period) or to patients whose EF was between 35% and 40%. In the Swedish population, MRA utilization was related to the experience of the provider. Patients who were followed by primary care were significantly less likely to receive MRAs than were patients who were being followed by specialists. Lower implementation of guideline recommended HF therapies by primary care providers than by specialists has been documented in other settings.14, 15 While this finding suggests that increased efforts at educating primary care providers about MRAs might improve their comfort in prescribing and maintaining these agents over time, the observation that MRA use was substantially lower than that for renin–angiotensin and sympathetic nervous system blockers in the Swedish HF Registry implies that there may be unique impediments to the use of MRAs. In the present report, MRA use did not decrease with elevated potassium levels but did with impaired renal function, even in the creatinine clearance 30–59.9 mL/min range. Based on this observation, the investigators suggest that perceived rather than actual risk of hyperkalaemia might have played a role in MRA non-use as neither hyperkalaemia nor worsening renal function appear to reduce the clinical benefit of these drugs.16 If this were the case, then approaches such as provider education or use of defined management pathways would be a good way of improving utilization. The reluctance of providers to use MRAs, however, may be more than perception. This registry report differs from clinical practice in that the investigators had access to lab values starting from the time of patient entry while the providers who were caring for the patients would have been able to see values obtained over the course of several years. Although patients were excluded from the registry analysis if their potassium or creatinine levels exceeded those defined in the guidelines, this was based on information collected around the time of inclusion into the registry. Provider knowledge of worsening of renal function or episodes of hyperkalaemia that preceded entry into the Swedish HF Registry, however, might have influenced the decision not to administer MRAs, was not available to the investigators, and could represent an unrecognized but legitimate reason for withholding an MRA. So, while we are left with a conundrum regarding exactly why MRAs are underutilized in HFrEF patients, the Swedish HF Registry results provide some targets for improving the situation. The low incidence of hyperkalaemia reported in the MRA clinical trials was likely due to both the stringent entry requirements in these trials and the carefully delineated recommendations to the investigators for avoiding hyperkalaemia during the course of the trials. As a result, the incidence of hyperkalaemia reported in the trials almost certainly underestimates what is seen in less well-regulated practice settings. Thus, the perception of hyperkalaemia with MRAs is deeply ingrained based on reports in the literature17-19 that may have been bolstered by the experience of individual providers with their own patients. This perception coupled with the evidence that MRAs are less likely to be prescribed in the primary care setting suggests that efforts focused on improving provider education by defining who is a candidate for an MRA, how to best administer MRAs, and strategies for avoiding hyperkalaemia are badly needed. Initiation of practice improvement pathways has also been shown to substantially increase implementation of guideline-directed HF therapies.20 The schism between primary care and specialty providers also suggests that societies like the European Society of Cardiology Heart Failure Association and the Heart Failure Society of America need to do a better job in reaching beyond their immediate membership to educate primary care providers, particularly since most HF patients are seen in a primary care setting. While some of the problems related to hyperkalaemia may be perceptual, there is a real possibility of a HFrEF patient becoming hyperkalaemic with the initiation of an MRA or while on maintenance therapy. Novel agents designed to specifically bind potassium in the gut and lower serum levels have either recently released or are being developed.21, 22 Their availability should enable more widespread use of drugs such as MRAs that are associated with hyperkalaemia. Although currently available clinical trial data suggest that both the continued use and higher dose of renin–angiotensin system blocking drugs including MRAs can be achieved with the use of these binding agents, further study in the HF population is needed in order for potassium binding agents to be more widely recommended in HFrEF patients who have experienced or at risk of developing hyperkalaemia. Also, MRAs that are less likely to produce hyperkalaemia are being evaluated in clinical trials and may offer an alternative to the currently available agents.23 The report by Savarese et al. in this issue of the Journal is an important reminder that although MRAs have been available for some time, there is still a lot of work to be done to educate providers about how and when to use them. As the gap between suboptimal implementation of MRAs and the use of most other evidence-based agents for treating HFrEF patients has persisted now for many years, the results of this study should serve as a wake-up call to individual providers, healthcare systems and professional societies alerting them to the fact that our HF patients cannot realize the benefits offered by MRAs unless they indeed take them. Conflict of interest: none declared.