Description of the condition
Heart failure is a leading cause of cardiovascular morbidity and mortality in the world. Due to the increased incidence and survival of people with cardiovascular diseases, heart failure has been singled out as an emerging epidemic (Braunwald 1997). A newly published report revealed that in 2009 there were over four million heart failure-associated hospitalizations in the United States (Blecker 2013).
Heart failure can be defined as a syndrome caused by an abnormality of cardiac structure or function leading to failure of the heart to pump blood at a rate commensurate with the requirements of the metabolizing tissues, despite increased filling pressures (McMurray 2012; McKelvie 2013). Whether heart failure results from an abnormality of cardiac structure or function, or both, it is characterized by symptoms such as fluid retention, shortness of breath, and fatigue, especially on exertion. Patients with heart failure also have characteristic signs such as elevated jugular venous pressure, pulmonary crackles, and peripheral edema. The prognosis for heart failure is poor, with reported survival estimates of 50% and 10% at 5 and 10 years, respectively (Cowie 2000; MacIntyre 2000).
Description of the intervention
Expanding knowledge of nuclear receptors has linked the essential role of glucocorticoid receptors (GRs) to homeostasis restoration (De Kloet 1998). In the classic model of nuclear receptor action, GR acts as a ligand-dependent transcription factor by either activating or repressing gene expression through direct interactions with DNA or other transcription factors (Labeur 2010). Evidence from animal studies suggests glucocorticoids may play an important role in body fluid homeostasis through regulating the production of atrial natriuretic peptide (ANP) and the expression of its primary receptor, natriuretic peptide receptor A (NPR-A) (Gardner 1988; Dananberg 1992; Liu 2010; Liu 2011). Glucocorticoids can act centrally as well as peripherally to assist in the normalization of extracellular fluid volume (Liu 2010; Liu 2011). Newly conducted clinical trials showed that glucocorticoids could potentiate renal responsiveness to diuretics in patients with heart failure and produce potent diuretic effects (Ashbel' 1985; Liu 2007; Liu 2013).
How the intervention might work
Newly emerging evidence shows that glucocorticoids not only have cardioprotective effects but also have renal protective effects. The cardioprotective effects of glucocorticoids are mediated by multiple mechanisms which include mediating nitric oxide production in the cardiac endothelium (Hafezi-Moghadam 2002) and up-regulating β-adrenergic receptor expression in the cardiac tissue (Nishimura 1997). The renal protective effect of glucocorticoids is mediated by increasing NPR-A expression in the kidneys (Liu 2011). Natriuretic peptide (NP) secretion, in response to volume expansion, provides an important mechanism to protect against salt and water retention (Potter 2006). NP levels elevate early on in heart failure. Despite a dramatic elevation of NP levels in decompensated heart failure, they fail to reverse the salt and water retention (McMurray 2012; McKelvie 2013). The down-regulation of expression of NPR-A, the primary receptor for the diuretic action of NP, in the kidney is thought to be one of the underlying mechanisms of NP resistance in heart failure (Schrier 1999; Bryan 2007). Glucocorticoids can up-regulate the expression of NPR-A in the kidney and hypothalamus (Liu 2011). Therefore, glucocorticoids could assist in the normalization of extracellular fluid volume through potentiating the actions of NPs in the hypothalamus and kidney (Liu 2010; Liu 2011). Moreover, glucocorticoids could specifically dilate the renal vasculature through enhancing nitric oxide (NO) production or activating the endogenous renal dopaminergic system (Aguirre 1999) resulting in an increase in glomerular filtration rate (Baylis 1978; May 1995; Aguirre 1999). There is a widespread assumption that glucocorticoids can cause fluid retention due to their structural similarity to mineralocorticoids (de Denus 2013), and thus should not be used in heart failure (McMurray 2012). The assumption that glucocorticoids could cause fluid retention is not supported by the data from either animal or human studies (Whitworth 1989; Aguirre 1999; Liu 2010). More importantly, glucocorticoids have been used to treat refractory heart failure since the 1950s (RIEMER 1958; MICKERSON 1959; NEWMAN 1959; RADO 1959; Korochkin 1976; Ashbel' 1985; Liu 2007). Glucocorticoids successfully reversed refractory heart failure in most of the cases. Newly conducted clinical studies and observational data show that glucocorticoids can produce a potent diuresis and an improvement in renal function in patients with decompensated heart failure (Ashbel' 1985; Liu 2006; Liu 2007; Massari 2012; Liu 2013). Therefore, glucocorticoids may be drugs of choice for patients with decompensated heart failure (Shanoff 1969; Johnson 2010).
Why it is important to do this review
Most patients with decompensated heart failure are admitted with overt volume overload or pulmonary congestion. Currently available therapeutic options in decompensated heart failure have limitations in their efficacy or safety, or both. Diuretics are an essential component of current treatment in such patients (McMurray 2012; McKelvie 2013). However, they induce diuretic effects at the cost of activation of the renin-angiotensin-aldosterone system and worsening of renal function (Francis 1985; Hasselblad 2007; Maeder 2012). Intravenous nitroglycerin is limited by rapid development of tolerance (Elkayam 2004). Dobutamine, dopamine, and milrinone improve haemodynamics (Cohn 1982; Cuffe 2002; Bayram 2005) but the beneficial effects are offset by an increased risk of ventricular arrhythmias and post-discharge mortality (Cuffe 2002; Bayram 2005; Tacon 2012). Nesiritide, a recombinant form of the 32 amino acid human B-type natriuretic peptide (BNP), failed to show superiority over standard care in reducing the rate of recurrent heart failure hospitalization or death at 30 days in the latest and largest 'Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure' (ASCEND-HF) trial (O'Connor 2011). Levosimendan, a calcium sensitizer, could improve haemodynamic parameters when compared with placebo yet it also failed to provide survival benefit (Ribeiro 2010). Ultrafiltration was once proposed as an alternative to loop diuretics for the treatment of patients with decompensated heart failure. However, recently conducted studies do not support this notion (Bart 2012). Given the fact that current treatments may cause adverse effects and the newer drugs do not improve survival despite their pathophysiological benefits, glucocorticoids may be an alternative therapeutic strategy for patients with decompensated heart failure.
Moreover, the evidence regarding the use of glucocorticoids in heart failure is limited. Glucocorticoids also have multiple potential side effects in large doses. These side effects may include fluid retention, disturbances in glucose tolerance, susceptibility to infection, sleep disorder, hyperactivity, osteoporosis, and muscle wasting (de Denus 2013; Liu 2013). Currently there is no systematic review examining the evidence, so a detailed Cochrane review will enable us to establish the efficacy and safety of glucocorticoids as an alternative treatment for decompensated heart failure.
The objective of this systematic review is to evaluate the effects of glucocorticoids in people with decompensated heart failure compared to standard medical treatments on mortality, renal function, New York Heart Association (NYHA) functional class, and their potential adverse effects (that is fluid retention).
Criteria for considering studies for this review
Types of studies
We will only include randomised clinical trials (RCTs) but we will exclude cross-over and cluster RCTs.
Types of participants
We will include trials involving adult participants (≧ 18 years) with decompensated heart failure who required medical treatment for their condition. The diagnosis of decompensated heart failure should be based on clinical symptoms with a combination of clinical signs or cardiac imaging and ideally the levels of BNP or N-terminal pro-brain natriuretic peptide (NT-proBNP).
The participants should present with at least one symptom of heart failure (dyspnoea at rest or with minimal exertion, orthopnoea), one or more accompanying signs of congestion (elevated jugular venous pressure, rales, peripheral edema, or ascites), and ideally one or more objective measures of heart failure (evidence of congestion or edema on chest radiography, elevated concentrations of BNP or NT-proBNP, pulmonary capillary wedge pressure > 20 mm Hg). There will be no prespecified inclusion criterion with respect to left ventricular ejection fraction (LVEF).
Exclusion criteria will include participants with decompensated heart failure complicated by myocarditis, inflammatory cardiomyopathy, or active rheumatic heart disease. Participants with pulmonary diseases such as asthma and pulmonary fibrosis, who may benefit from glucocorticoid therapy, will also be excluded.
Types of interventions
Intervention: any glucocorticoid (given orally or intravenously) in combination with standard care.
Comparison: a placebo in combination with standard care, or standard care alone.
Types of outcome measures
- Short-term mortality (≦ 30 days)
- Medium-term mortality (one month to six months)
- Long-term mortality (longer than six months)
- Change from baseline in serum creatinine (SCr)
- New York Heart Association (NYHA) functional class
- Change in volume status, i.e. occurrence of fluid retention expressed by percentage of participants, or change from baseline in body weight
- Adverse events, such as hyperglycaemia requiring insulin infusion, all causes of infection, sleep disorder, hyperactivity, osteoporosis, and muscle wasting
Search methods for identification of studies
The literature searches will be undertaken by the Cochrane Heart Group Trials Search Co-ordinator using the following sources: the Cochrane Central Register of Controlled Trials (CENTRAL) in The Cochrane Library, which contains the Cochrane Heart Group Trials Register; MEDLINE (Ovid), and EMBASE (Ovid).
The preliminary search strategy, designed for MEDLINE, can be found in Appendix 1. This strategy will be adapted for use in the other databases. We will apply RCT search filters which are based on the Cochrane sensitivity-maximising RCT filter for MEDLINE (Ovid) (Lefebvre 2011).
We will not apply any language restrictions.
Searching other resources
We will search two clinical trial registers, the World Health Organization International Clinical Trials Registry Platform (ICTRP) (http://apps.who.int/trialsearch/) and ClinicalTrials.gov. We will contact the authors of included studies for potentially missing data and to identify other ongoing or completed trials.
We will also search SinoMed (Chinese) to retrieve relevant articles.
We will handsearch the reference lists of retrieved articles for additional trials.
Data collection and analysis
Selection of studies
Two authors (NW, and HM) will independently select trials for inclusion in the review. We will discuss the differences and refer to the third author (MQ or CL) where differences persist. We will contact the authors of studies if we need more information to determine eligibility. We will exclude all irrelevant titles and obtain full-text papers when we think the titles or abstracts are relevant or their eligibility is unclear. Two authors (NW and CL) will assess these full-text papers and we will resolve disagreements by bringing in the third author (MQ). We will translate non-English articles and describe them in the 'Characteristics of included studies' table. We will use the reference management software Endnote to manage the records retrieved from the searches of the electronic databases.
Data extraction and management
Three authors (NW, MH, and CL) will conduct data extraction, and we will resolve disagreements by discussion or referring to the fourth author (MQ). We will extract the study characteristics and outcome data in duplicate. We will use the data extraction form provided by the Cochrane Handbook for Systematic Reviews of Interventions to obtain the data and ensure that the authors retrieve comparable results. We will include the data falling into the following categories.
- Mortality: short-term mortality, medium-term mortality, and long-term mortality.
- Change from baseline in SCr.
- Change in NYHA functional class.
- Change in volume status, i.e. occurrence of fluid retention expressed by percentage of participants, or change from baseline in body weight.
- Hyperglycaemia requiring insulin infusion.
We will collect the information, such as context, implementation, cost, and sustainability, and report the data in the 'Characteristics of included studies' table. We will include all potential moderators or confounders of the study outcomes in the extraction form. If data on multiple measures of the same or similar outcomes are available in a single study, we will only pool the data we specified in the protocol and describe the other measures of the same or similar outcomes in the supplementary tables.
Assessment of risk of bias in included studies
We will independently assess the methodological quality of the included trials against the following criteria (Higgins 2011).
- Was the study described as randomised?
- Was the allocation concealment adequate?
- Were the participants blinded?
- Was there blinded outcome assessment?
- Was there a description of withdrawals and dropouts?
- Were the results analysed according to intention to treat?
Measures of treatment effect
We will combine results using an odds ratio (OR) for dichotomous data and mean difference (MD) for continuous data. We will treat short ordinal scales (for example NYHA class) as dichotomous measures (we will arrange the data into two categories: improvement or deterioration in class).
Unit of analysis issues
For trials having more than two treatments, we will combine treatment data and compare them collectively with the control group (Higgins 2011).
Regarding repeated measurements, we will only analyse the outcomes recorded at our prespecified time points, that is short-term, medium-term, and long-term.
Dealing with missing data
For missing data on prespecified primary outcomes, we will contact the study investigators to request the data. If the data are not available, we will perform sensitivity analysis to assess how sensitive the results are to reasonable changes in the assumptions that we make.
Assessment of heterogeneity
We will compare the Chi
Assessment of reporting biases
We will conduct standard funnel plots for the primary outcome, that is mortality, to investigate the potential for publication bias influencing the results.
In the absence of significant heterogeneity, we will combine the results with a fixed-effect model using ORs for dichotomous data and MDs for continuous data (Higgins 2011). If there is significant heterogeneity (that is the ratio of the Chi
Subgroup analysis and investigation of heterogeneity
- Patients with acute decompensated heart failure (patients hospitalized for decompensated heart failure, occurring within 24 hours) versus patients with chronic decompensated heart failure (patients hospitalized for decompensated heart failure, occurring at longer than 24 hours)
- Decompensated heart failure patients with left ventricular systolic dysfunction (LVEF) < 45% versus patients with preserved LVEF ≧ 45%
- Low-dose (≦ 7.5 mg prednisone equivalent daily), medium-dose (> 7.5 mg but ≤ 100 mg prednisone equivalent daily) and high-dose (> 100 mg prednisone equivalent daily) systemic glucocorticoid therapy (Duru 2013)
When there is significant heterogeneity in the included studies, we will conduct a sensitivity analysis to test if including trials of lower methodological quality changes the effect estimates.
Appendix 1. MEDLINE search strategy
1 exp Glucocorticoids/
3 (viarin or beclo azu or beclazone or apo-beclomethasone or aldecin or propaderm or filair or bemedrex or ventolair or ascocortonyl or asmabec or aerobec or prolair or junik or becloturmant or becloforte or beclomet or beclocort or beclamet or sanasthmyl or becodisk or vancenase or beclovent or nasobec aqueous or aerobec forte or vanceril or becotide or beclorhinol or respocort or beconase or beclo asma or qvar or sanasthmax or bronchocort or beclazone or ecobec or becodisks).tw.
5 (cellestoderm or betadexamethasone or celeston* or flubenisolone).tw.
7 (pulmicort or horacort or rhinocort).tw.
9 (olux or embeline or temovate or clobex or cormax or clobetasol or clofenazon or dermovate).tw.
11 (deoxydexamethasone or a41304 or a 41304 or topicort or stiedex or topicorte or ibaril or a-41304 or desoxi or topisolon or flubason).tw.
13 (millicorten or maxidex or decaspray or dexpak or dexasone or oradexon or decaject* or hexadecadrol or hexadrol or methylfluorprednisolone or decameth).tw.
17 "fluocinolone acetonide".tw.
18 (jellin or synalar* or flucinar or alvadermo or capex or flusolgen or syn?mol or co-fluocin or "fluortriamcinolone acetonide" or derma-smooth or derma-smooth or jellisoft or fluonid or fluodermo or cortiespec or fluocid or "derma smooth").tw.
20 (lidex or lidemol or fluocinolide or topsym or novoter or tiamol or klariderm or lyderm or topsyn* or metosyn).tw.
24 (efflumidex or cortisdin or fluor-op or fml or "fluor op" or fluoro-ophtal or "fluoro ophtal" or flucon or fluoropos).tw.
27 (flurandrenolide or haelan or cordran).tw.
30 (medrol or metipred or urbason).tw.
33 (diadresonf or predate or predonine or "di adreson f" or di-adreson-f).tw.
35 (dehydrocortisone or decortin or encorton or predniment or kortancyl or enkortolon or decortisyl or rectodelt or meticorten or encortone or dacortin or "predni tablinen" or cortancyl or sone or panafcort or delta-cortisone or cortan or "liquid pred" or pronisone or orasone or winpred or panasol or prednidib or sterapred or cutason or deltasone or ultracorten).tw.
37 (aristocort or volon).tw.
38 (kenalog or tricort-40 or tricort40 or "tricort 40" or cinonide or azmacort or kenacort).tw.
39 (hydrocortisone or alclometasone or algestone or clobetasone or clocortolone or cloprednol or cortisone or corticosterone or deflazacort or desonide or desoximetasone or diflorasone or difluprednate or fluazacort or flucloronide or flunisolide or "fluocortin butyl" or "fluperolone acetate" or formocortal or halcinonide or "halobetasol propionate" or halometasone or "halopredone acetate" or hydrocortamate or "loteprednol etabonate" or medrysone or meprednisone or "mometasone furoate" or prednicarbate or prednylidene or rimexolone or 21-acetoxy-pregnenolone or cortivazol or amcinonide or "fluticasone propionate" or mazipredone or tixocortol).tw.
41 Heart Failure/
42 ((heart or cardiac or myocard*) adj2 failure*).tw.
43 (heart adj2 decomp*).tw.
45 40 and 44
46 randomized controlled trial.pt.
47 controlled clinical trial.pt.
50 drug therapy.fs.
54 46 or 47 or 48 or 49 or 50 or 51 or 52 or 53
55 exp animals/ not humans.sh.
56 54 not 55
57 45 and 56
Contributions of authors
NW, MQ, HM, and CL participated in the conception, design, and drafting of the protocol.
Declarations of interest
Sources of support
- Cochrane Heart Group, Not specified.The search strategies for databases was established by Cochrane Heart Group.
- This is part of a project sponsored by China Scholar Council (201208130092)., China.