Progress or lack of progress in hospitalized heart failure

Keywords::

• clinical update
• heart failure
• outcomes

Hospitalized heart failure (HHF) is a common occurrence in 2013 but has only recently been recognized as a clinical entity by clinicians, industry, regulators and payers [1]. Heart failure ranks as the most common reason for admission in elderly in the USA, accounting for extraordinarily high medical expenditures [2]. Heart failure patients after index hospitalization face approximately a 40% risk of death or rehospitalization within 6 months post-discharge [3]. This unacceptably high post-discharge event rate has not changed substantially or has only modestly improved over time, despite available drug- and device-based therapies. Thus, HHF may represent one of the most pressing issues in medicine today. Despite these dismal statistics, the last decade in HHF research has provided significant insights into our overall understanding of this complex condition. In this article, we have summarized the major lessons learned over the last decade to help guide the necessary steps moving forward in clinical and research settings.

Heterogeneity of HHF

Heart failure is no longer considered a single entity but rather as different cardiac and noncardiac conditions culminating into a common pathway. Historical approaches to HHF management have largely been empiric and generic. However, we have learned from global registries and trials that widespread differences are observed in initial clinical presentation, baseline characteristics, precipitating factors and comorbid diseases among HHF patients. The mean age of HHF patients is approximately 70 years with about 40% females and 20% blacks in the USA [4]. Coronary artery disease (CAD) and hypertension are the most common comorbidities, each present in up to 70% of patients. Other common comorbidities include renal insufficiency, diabetes mellitus, chronic lung disease and atrial fibrillation (present in 30–40% of patients). The majority of patients are hypertensive or normotensive at presentation, which may reflect underlying cardiac reserve. Dyspnea and not low cardiac output is the main reason for admission. Approximately, 50% of patients have relatively preserved ejection fraction (EF). In spite of marked improvement during hospitalization in signs and symptoms of HF, the majority of patients are discharged home in hemodynamic congestion, reflected by high levels of circulating natriuretic peptides. Over the last decade, younger patients and blacks have not experienced similar improvements in in-hospital outcomes as other segments of the population [5]. Site-based and region-specific variations has been noted in clinical trials of HHF [6,7], suggesting more systematic influences of HHF disease presentation, management and outcome. Indeed, lack of success in recent HHF trials may be due to broad inclusion criteria. New classification schemes have helped better define these subsets and may guide future tailored management strategies in HHF.

Prognosis determination

Three principle responsibilities of a physician are the diagnosis, management and prognostication. Prior research has focused almost exclusively on the former two roles; however, prognosis has been more recently recognized as a key area. A number of clinical risk models for in-hospital and post- discharge mortality have been derived and validated for HHF patients [8,9]. However, the extent to which these risk models are applied in routine clinical practice is not well defined. Hyponatremia occurs in approximately 25% of all admissions and does not change appreciably during hospitalization. Although low serum sodium predicts poor post-discharge event rates [10], management of HHF patients with a vasopressin antagonist does not appear to influence clinical end points [11]. Low systolic blood pressure at the time of admission and after initial stabilization is one of the strongest determinants of inpatient and post-discharge mortality [12]. Post-discharge readmission rates appear to be similar, regardless of blood pressure [12]. Abnormal baseline renal function and in-hospital worsening renal parameters have emerged as major prognostic factors, perhaps due to limitation of use of traditional HF therapies [13]. Approximately, a quarter of patients experience worsening renal function during hospitalization for HF, likely secondary to underfilling, rather than worsening HF and venous congestion [13]. In addition, widened QRS duration, suggesting underlying dyssynchrony, is evident in roughly 40% of HHF patients with reduced EF and is a marker of high post-discharge event rates [14]. Although cardiac resynchronization therapy appears to improve outcomes in ambulatory patients with widened QRS durations, its effect in this patient population remains to be determined. Cardiac [15] and noncardiac [16] comorbidities inform risk of post-discharge events. It is unclear whether these identified parameters are markers of disease severity or mediators of poor outcomes. Regardless, these high-risk subgroups may be the focus post-discharge monitoring programs.

Redefining end points

Dyspnea has been and continues to be a major end point in the evaluation of new therapies for inpatient management of HHF. However, it has become apparent that dyspnea is readily and rapidly improved by standard therapies [17]. In addition, dyspnea is subjective and often inconsistently documented, despite efforts to standardize measurement. More recent large Phase III clinical trials have included definitive end points such as post-discharge mortality and rehospitalization [3]. Readmission after index HHF has captured the attention of payers, policy makers, patients and physicians, given the large financial and clinical burden of these events. Readmission has emerged as an end point for clinical trials, benchmark for quality and a guide for reimbursement.

Existing & novel therapeutics

HF represents a challenging substrate for pharmacologic and device interventions. Unlike acute coronary syndromes where the target (clot and coronary occlusion) and intervention (revascularization) are linear and logical, such relationships are not as consistent in HF. For example, interventions aimed at augmenting pump function and contractility in patients with severe systolic dysfunction paradoxically worsen mortality, due to attendant risks of myocardial damage and arrhythmias.

HHF clinical trials can be broadly classified as stage A (early, short-term interventions), B (in-hospital therapies) or C (therapies continued into post-discharge period) [18]. Traditionally, pivotal HHF trials have tested novel vasoactive agents (stage A) as a bridge to recovery during acute decompensation, yet, barring one exception, every trial conducted to date has been negative in terms of efficacy and/or safety. Short-term infusion therapies of vasoactive agents cannot mediate long-term outcomes unless significant end-organ damage is prevented. Newer trials should potentially evaluate therapies instituted after initial stabilization of HHF patients and continued post-discharge (stage B and C).

Aggressive inpatient diuresis continues to be the mainstay of therapy for HHF. It is clear that many patients are insufficiently decongested and thus, end points for inpatient diuresis require further clarification. It should be recognized that diuresis remains palliative and has not been definitely demonstrated to alter HF clinical course. Recent data suggest that there are no significant differences in symptoms and renal function with bolus versus continuous infusion or high- versus low-dose diuretic therapy [19]. Furthermore, other methods of volume removal including ultrafiltration have not been shown to improve clinical outcomes compared to standard step-up diuresis in patients with cardiorenal dysfunction [20].

While intravenous inotropic agents such as dobutamine and milrinone were often administered to HHF patients even in the absence of hypoperfusion, data from the last decade from clinical trials and observational studies have consistently demonstrated that these agents increase serious adverse events and mortality particularly in patients with CAD [21–23]. The concomitant use of β-blockers does not appear to attenuate this excess risk [24]. Based on this cumulative data, the utilization of inotropes other than digoxin in the management of HHF is relatively contraindicated and may contribute to the progression of HF even when they are used only for few days [25]. Although recognition of their deleterious effects has led to a decrease in intravenous inotrope use in HHF [26], a significant number of patients continued to receive this therapy. This practice should be strongly discouraged.

Initiation and continuation of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers and β-blockers in the absence of contraindication has been shown to improve post-discharge outcomes for HHF patients with reduced EF [27,28]. As a result of quality improvement registries, there has been a marked increase in use of these guideline-directed medical therapies in HHF patients [29]. The use of mineralocorticoid antagonists in HHF has been tempered, likely secondary to risks of hyperkalemia. Although these agents have not been tested in this setting, data from patients with post-myocardial infarction left ventricular dysfunction are promising [30]. New mineralocorticoid antagonists are under development that may have more tolerable safety profiles [31]. The use of digoxin has not been tested in HHF; however, the drug appears to safely improve hemodynamic profiles outcomes in chronic HF, especially those with severe signs/symptoms of disease [32]. Digoxin significantly reduces 30-day all-cause hospitalization rates in ambulatory HF patients [33]. Despite this, the rates of digoxin use and uptake have greatly diminished in the last decade.

Nesiritide, a synthetic B-type natriuretic peptide, initially approved in 2002 for HHF, appears to improve hemodynamics and has a minor effect on dyspnea. In a large, global clinical trial of over 7,000 patients, nesiritide was found to be safe but not effective in improving clinical outcomes [34]. Rolofylline, an adenosine A-1 receptor antagonist, was intended to improve renal function and thus clinical outcomes in HHF. However, it similarly did not meet expectations in a large Phase III clinical trial. No differences were noted between the rolofylline arm and placebo in terms of 60-day death or readmission and persistent renal impairment [35]. Intravenous levosimendan, a calcium sensitizer, showed promise in achieving short-term stabilization of clinical status in HHF up to 5 days into hospitalization, at the expense of excess hypotension, ventricular tachycardia and atrial fibrillation compared to placebo [36]. There was also a trend on increase in mortality, particularly in patients with CAD, and in whom the drug caused hypotension [36]. No differences in clinical outcomes were noted between levosimendan and dobutamine [37]. Vasopressin antagonism is known to both induce aquaresis and improve or normalize hypernatremia. Tolvaptan, an oral vasopressin-2 receptor antagonist, was associated with a significant decrease in body weight and dyspnea during hospitalization without influencing renal function. In spite of these short-term benefits, the long-term use of tolvaptan did not improve post-discharge outcomes [38]. The clinical effects of tolvaptan in HHF patients with hyponatremia are presently unclear. The use of intravenous cGMP activators was associated with an improvement in hemodynamics in HHF, but was associated with higher rates of hypotension. Accordingly, short-term use was no longer investigated [39]. The chronic use of this agent is currently under active investigation and may reveal more favorable results [40]. The addition of a renin inhibitor (aliskiren) to standard therapy had no effect in overall post-discharge mortality and rehospitalization but was associated with a marked decrease in a number of biomarkers including troponin, aldosterone and natriuretic peptides [3]. Patients without diabetes may derive a mortality benefit at 12 months from aliskiren, but this hypothesis remains to be formally tested [3]. Most recently, recombinant relaxin, a natural vasodilator, produced during pregnancy, was shown to significantly improve dyspnea in HHF [41]. It is interesting that its short-term use of serelaxin was associated with a marked reduction in mortality at 180 days. The available data suggest that this drug in addition to being a vasodilator may prevent cardiac and renal injury during hospitalization and in effect, prevent progression of HF [41]. The hypothesis remains to be tested and is an area of active interest in HHF. As heart rate is increasingly becoming recognized as an important target in HHF, the selective sinus node inhibitor ivabradine, which has demonstrated initial success in ambulatory patients [42], may have a role in the early post-discharge period.

Hospitalization may be the ideal time-point for reconciliation and early initiation of therapies known to be effective in chronic HF. Our understanding of the underlying pathophysiology of HF must evolve and mature to formulate effective interventions directed at modifiable targets. We must identify specific clinical phenotypes within HHF and tailor therapies accordingly. We have made great strides in uncovering new insight into our overall approach to this high-risk population; however, the drug and device armamentarium for this syndrome is limited. Hospitalized heart failure stands as a unique syndrome, enriched with high-risk patients that have failed outpatient management and may benefit from inpatient and post-discharge therapeutics. The next decade will be dedicated to translating this understanding to much needed novel therapies in HHF.

Acknowledgements

The authors would like to thank Fumiko Inoue for her excellent technical assistance with the preparation of this manuscript.

Financial & competing interests disclosure

J Butler is a consultant for Stemedica, Medtronic, Bayer, Novartis, Amgen, Gambro, Harvest, Celladon, Ono, and Trevena. GC Fonarow has received research grants from NHLBI, AHRQ and is a consultant for Novartis, Gambro and Medtronic. M Gheorghiade is a consultant for Abbott Laboratories, Astellas, Astra Zeneca, Bayer Schering Pharma AG, CorThera, Inc, Cytokinetics, Inc, Debiopharm SA, Errekappa Terapeutici (Milan, Italy), GlaxoSmithKline, Johnson & Johnson, Medtronic, Merck, Novartis Pharma AG, Otsuka Pharmaceuticals, Pericor Therapeutics, Protein Design Laboratories, Sanofi-Aventis, Sigma Tau and Solvay Pharmaceuticals. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.