7. Treatment

7.1 Chronic HF

Pharmacotherapy has been shown to change the natural history of HFrEF. HFpEF however, has been identified as major public health issue and to date, the etiology, diagnosis, characterization, and treatment has remained challenging. Goals of HF therapy include improving survival and reducing morbidity such as hospitalizations and symptoms, while improving functional capacity and quality of life. Figure 4 shows a therapeutic approach to patients with HFrEF that is considered optimal medical therapy and defined as GDMT throughout this section. The evidence-based medications and doses of GDMT are shown in Table 11.

7.1.1 HFrEF pharmacological treatment

Contemporary treatment for most patients with HFrEF encompasses triple therapy, which includes the combination of: (1) an ACEi (or ARB if ACEi-intolerant); (2) a β-blocker; and (3) an MRA. Working on various pathways of the neurohormonal system, the combination of these agents has been shown to improve survival in patients with HFrEF. There are many landmark trials and meta-analyses that support the use of ACEis^[93]–[99] and β-blockers^{[100]–[104]} in all patients across the spectrum of HFrEF. ARBs have been shown to be superior to placebo in those intolerant to ACEis and are considered a good second-line agent.^{[105]–[108]} Likewise, there are 2 key clinical trials and 1 meta-analysis^{[109]–[111]} that support the additional use of an MRA with this combination with an improvement in survival across the spectrum of symptomatic patients with HFrEF. Most recently, the Eplerenone in Mild Patients Hospitalization and Survival Study in Heart Failure (EMPHASIS-HF) expanded the use of aldosterone receptor antagonists in HFrEF patients with mild symptoms.^[110] In EMPHASIS-HF the effects of eplerenone on clinical outcomes were examined in patients 55 years of age or older, with NYHA II symptoms, LVEF < 30% (if > 30%-35%, a QRS duration of > 130 ms), treated with an ACEi and/or ARB, and β-blockers. A total of 2737 patients with a median follow-up of 21 months were enrolled. There was a 37% reduction in the primary composite outcome of death from cardiovascular causes or first hospitalization for HF with eplerenone.

7.1.1.1 Pharmacologic therapy

In addition to initiating, titrating, and monitoring pharmacologic therapy, there are circumstances in which some therapies may be withdrawn (Table 12). There are additionally some common effects of GDMT requiring active surveillance and management. A suggested approach to hyperkalemia is presented in Table 13.

7.1.1.2 ACEi/ARB

There are extensive data on the use of ACEi and β-blocker treatment for patients with HFrEF to reduce morbidity and mortality and improve quality of life.^[112],[113] A notable deletion from these guidelines is the recommendation to consider combination ACEi and ARB therapy, previously recommended. The combination of an ACEi with an ARB is no longer recommended. Although some evidence exists to support a reduction in clinically relevant outcomes with the combination, there is also substantial evidence that was published after the previous recommendation, outlining harm in terms of adverse effects (eg, hypotension, hyperkalemia, and renal dysfunction).^{[108],[114],[115]} More contemporary treatments with MRAs and ARNIs have a stronger evidence base across the spectrum of outcomes (eg, morbidity and mortality) and therefore further limit the role of combination ACEi and ARB therapy.

7.1.1.3 β-Adrenergic receptor blocker (β-blocker)

β-Blockers are part of the first-line therapy in the treatment of HFrEF, because they have been proven to improve survival and decrease hospitalizations in this population of patients, in a number of large clinical trials.^{[101],[103],[116]-[121]}

7.1.1.4 MRAs

A single RCT supports the use of eplerenone (target 50 mg daily) compared with placebo post-MI.^[122] The Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study (EPHESUS) trial enrolled 6642 patients who had an MI 3-14 days previously with an LVEF < 40% and symptoms of HF or an LVEF < 30% and diabetes without symptoms of HF. The primary outcome included all-cause mortality and cardiovascular mortality or hospitalization for cardiovascular events. After a median follow-up of 16 months, there was a 15% relative decrease in mortality and 13% relative decrease in cardiovascular mortality or hospitalization for cardiovascular events in the eplerenone group. There was more hyperkalemia in the eplerenone group.

7.1.1.5 ARNI

In those who remain symptomatic despite triple therapy, consideration should be made to change an ACEi/ARB to an ARNI. Neprilysin, a neutral endopeptidase, degrades several endogenous vasoactive peptides, including NPs, bradykinin, and adrenomedullin. Inhibition of neprilysin increases the levels of these substances, countering the neurohormonal overactivation that contributes to vasoconstriction, sodium retention, and maladaptive remodelling.^[123],[124] In the Prospective Comparison of ARNi With ACEi to Determine Impact on Global Mortality and Morbidity in Heart Failure (PARADIGM-HF) trial, the ARNI sacubitril/valsartan was compared with enalapril in patients with HFrEF.^[125] A total of 8442 patients were randomized to sacubitril/valsartan 200 mg twice daily or enalapril 10 mg twice daily after a 6-8 week runin phase. Patients were included if they were NYHA class II-IV (70% class II), LVEF ≤ 40% (amended to ≤ 35%), had a BNP ≥ 150 pg/mL (or NT-proBNP ≥ 600 pg/mL), or hospitalization for HF in the past year and BNP ≥ 100 pg/mL (or NT-proBNP ≥ 400 pg/mL). The primary outcome was a composite of death from cardiovascular causes or hospitalization for HF. The trial was stopped early, according to prespecified rules, after a median follow-up of 27 months. The primary outcome occurred in 914 patients (21.8%) in the sacubitril/valsartan group and 1117 patients (26.5%) in the enalapril group, a 20% relative reduction. There was also a decrease in all-cause mortality, cardiovascular mortality, HF hospitalization, and symptoms of HF. The sacubitril/valsartan group had a higher proportion of patients with hypotension but a smaller risk of renal impairment, hyperkalemia, and cough than the enalapril group. The type of patients and magnitude of effect were similar to other landmark trials in HFrEF including ACEis, β-blockers, and MRAs. This trial also closely reflects contemporary practice with high utilization of ACEis, β-blockers, and MRAs (100%, 92%, and 55%, respectively) at baseline and had an active gold standard comparator.

7.1.1.6 Ivabradine

Resting heart rate independently predicts CVD events, including HF hospitalization.^[126],[127] Systematic reviews have shown that a major contributor to the benefits of b-blocker therapy might be their rate-lowering effect.^{[128]–[130]} Despite their benefits, β-blockers are generally underused and underdosed.^{[129]–[131]} Ivabradine is approved for the treatment of HF by Health Canada. The latter drug selectively inhibits the depolarizing If current in the sinus node. It thus requires sinus rhythm to provide its pharmacological effect. In contrast to β-blockers, ivabradine does so without lowering BP or myocardial contractility.^[132],[133]
The first trial to assess ivabradine in CAD was the Morbidity-Mortality Evaluation of the I_f Inhibitor Ivabradine in Patients With Coronary Disease and Left-Ventricular Dysfunction (BEAUTIFUL) trial.^[134] In this trial the effect of ivabradine 7.5 mg twice daily was evaluated in patients with CAD and LVEF < 40% in sinus rhythm with a heart rate > 60 bpm in > 10,000 patients. Although ivabradine did not reduce the primary composite end point of cardiovascular death, hospitalization for MI, or new-onset or worsening HF, it did reduce the incidence of the secondary end point of fatal and nonfatal MI in patients with a baseline heart rate ≥ 70 bpm.
The Systolic Heart Failure Treatment With the I_f Inhibitor Ivabradine Trial (SHIFT) trial was the key trial to address the use of ivabradine in symptomatic HF.^[135] Inclusion criteria were NYHA class II-IV, sinus rhythm, resting heart rate ≥ 70 bpm, LVEF ≤ 35%, and HF admission within 12 months. Patients were randomized to a target dose of ivabradine 7.5 mg twice daily vs placebo. The primary end point was a composite of cardiovascular death or HF admission. Ninety percent of patients were receiving a β-blocker, and 56% were receiving 50% of target doses. Heart rate was 8 bpm lower in the ivabradine group at the end of the study. There was an 18% decrease in the primary outcome, which was largely driven by hospital admission for worsening HF (RRR, 26%). Treatment effect was consistent across prespecified subgroups, although the difference between treatment groups did not reach statistical significance in the subgroup with a baseline heart rate lower than the median of 77 bpm. Additionally, in those receiving > 50% of the target dose of a β-blocker, the overall trial results were similar. Ivabradine did not reduce all-cause or cardiovascular mortality. There were more withdrawals (21% vs 19%) and bradycardia in the ivabradine group (10% vs 2%). Only 1% of patients withdrew from the study as a consequence of bradycardia. Visual symptoms specific to ivabradine occurred rarely (3% vs 1% with placebo; P < 0.0001 and led to withdrawal in 1% of cases).

7.1.1.7 Hydralazine and isosorbide dinitrate

Three RCTs inform the use of H-ISDN in HFrEF. The Vasodilator in Heart Failure Trial (V-HeFT) trial, the first RCT, compared the effect of H-ISDN, prazosin and placebo in HFrEF on mortality (n = 642).^[136] After a mean follow-up of 2.3 years, there was no difference in mortality for the entire follow-up period (primary outcome), but showed a 66% relative improvement in survival in the H-ISDN group at 2 years. This trial predated the era of ACEis and β-blockers. The second trial to evaluate H-ISDN (300 mg and 160 mg) compared with enalapril (20 mg daily) in HFrEF on the outcome of mortality (n = 804).^[137] There was a reduction in mortality in the enalapril arm after a mean of 2.5 years (32.8% vs 38.2%; P = 0.016) and no difference in hospitalizations. Neither of these trials provide an insight into the role of H-ISDN in the face of contemporary therapy. The third trial was the African-American Heart Failure Trial (A-HeFT) trial, in which H-ISDN was investigated in additional to optimal therapy (ACEi/ARB, β-blocker, MRA) in self-identified black patients with NYHA class III/IV HFrEF.^[138] Black patients were specifically evaluated in this trial because it had been noted that this population has a less active renin-angiotensin system and seemed to respond better to H-ISDN. In this trial H-ISDN (225 mg or 120 mg) was evaluated vs placebo (in addition to standard therapy) on the outcome of all-cause mortality, first hospitalization for HF, and quality of life. A total of 1050 black patients were enrolled and followed for a mean of 10 months. The study was terminated early secondary to higher mortality in the placebo group. The primary outcome was a weighted score, but individual components of the outcome showed a difference favouring H-ISDN for all-cause mortality, first hospitalization for HF, and change in quality of life score. It is unclear if these results can be extrapolated to other groups.

7.1.1.8 Digoxin

The effect of digoxin on mortality and morbidity in patient with heart failure (Digitalis Investigation Group [DIGtrial])^[139] enrolled 6800 patients with HF and a LVEF ≤ 45% and were randomized to digoxin (median dose 0.25 mg/d) or placebo. The primary outcome was mortality over a mean follow-up of 37 months. Fifty-four percent were NYHA class II and 94% of patients were receiving an ACEi. There was no difference in all-cause mortality. There was a decrease in HFrelated deaths but an increase in “other cardiac deaths,” which has led to speculation that it might be due to arrhythmic death and led to an overall neutral effect on mortality. There were fewer patients hospitalized for HF in the digoxin group. Suspected digoxin toxicity was higher in the digoxin group (11.9% vs 7.9%). A systematic review included 13 studies (n = 7896, 88% of participants from the DIG-trial) showed similar results.^[140] None of these studies provide much insight into the relative benefit or harm of digoxin in light of contemporary therapy with β-blockers and MRAs, however, many landmark trials of these agents had a substantial background therapy of digoxin with no apparent change in the overall results if a patient was or was not receiving digoxin.

7.1.1.9 Omega-3 polyunsaturated fatty acid

The Gruppo Italiano per lo Studio della Sopravvivenza nell’Insufficienza cardiaca-Heart Failure (GISSI-HF) study was an RCT designed to assess the effects of omega-3 polyunsaturated fatty acids (n-3 PUFAs) in HF.^[141] More than 4600 patients with NYHA class II to IV HF, irrespective of etiology or EF, were randomly assigned to a fish-based n-3 PUFA (daily 850 mg to 882 mg eicosapentaenoic acid and docosahexaenoic acid as ethyl esters in the average ratio of 1:1.2) or placebo. The primary end points were time to death, and time to death or admission to hospital for cardiovascular reasons. After a median 3.9-year follow-up, there was a decrease in both primary outcomes favouring n-3 PUFA (9% relative reduction in all-cause death and an 8% relative reduction in death or admission to hospital). The therapy was well tolerated with primarily gastrointestinal side effects, and fewer than 10% of patients required study drug withdrawal. Current sources of n-3 PUFA in Canada are food supplements, therefore, are not subject to the regulatory review (including predefined tolerances for drug content) that is required for any drug approval. As such, it is difficult to be certain of the amount of n-3 PUFA present in any given commercial preparation. Indeed, evidence suggests a large degree of variability between different available forms of n-3 PUFA.^[142] Patients and caregivers who wish to use n-3 PUFA are therefore referred to a local medical practitioner, pharmacy, or other reputable source of information to determine their best source of n-3 PUFA. Reports of excessive bleeding have been associated with doses < 3 g/d, but this remains controversial.^[143],[144]

7.1.1.10 3-Hydroxy-3-methyl-glutaryl-coenzyme A reductase inhibitors (statins)

Many patients with HF have coexistent ischemic heart disease; however, these patients were systematically excluded from many of the early landmark statin trials. Two RCTs give insight into the benefit of statins specifically in patient with HF.
The Controlled Rosuvastatin Multinational Trial in Heart Failure (CORONA) study was an RCT of 5011 patients with HF that compared rosuvastatin 10 mg/d with placebo.^[145] There was no difference in the primary end point of cardiovascular mortality, nonfatal MI, or nonfatal stroke. There was an 8% relative reduction in the secondary outcome of cardiovascular hospitalizations, but not HF hospitalizations. Rosuvastatin was well tolerated, with fewer withdrawals from therapy than with placebo. Despite achieving the expected low-density lipoprotein cholesterollowering of rosuvastatin, there was little benefit in this cohort of patients with CAD.^[146]
The second trial was the GISSI-HF study; 4574 patients with chronic HF, NYHA class II-IV, irrespective of cause and LVEF, were randomly assigned to rosuvastatin 10 mg/d or placebo, and followed for a median of 3.9 years.^[147] There was no difference in the primary end points of time to death, and time to death or admission to hospital for cardiovascular reasons. There was no difference in any other outcomes or subgroups.

7.1.1.11 Anticoagulation and antiplatelet therapy

There are no RCTs that evaluate the role of ASA in comparison with placebo in patients with HF. A meta-analysis showed a reduction in serious vascular events, stroke, and coronary events with ASA therapy in secondary prevention trials.^[148]
The 2 largest RCTs both compared warfarin with ASA (with or without clopidogrel) rather than placebo. The Warfarin and Antiplatelet Therapy in Chronic Heart Failure (WATCH) trial compared warfarin (INR, 2.5-3), ASA (162 mg) and clopidogrel (75 mg) in patients with HFrEF in sinus rhythm. A total of 1587 patients were followed for a mean of 1.9 years.^[149] The study was stopped early secondary to poor recruitment. There was no difference in the primary end point of all-cause mortality, nonfatal MI, or nonfatal stroke in any of the groups. However, there was a reduction in stroke in the warfarin arm compared with the antiplatelet arms, but there was also a higher risk of bleeding in the warfarin group compared with the clopidogrel group. The Warfarin versus Aspirin in Reduced Cardiac Ejection Fraction (WARCEF) trial, the largest trial to date, had similar results with a single comparison arm of ASA 325 mg daily vs warfarin (INR, 2-3.5). A total of 2305 patients were enrolled with a mean follow-up of 42 months.^[150] There was no difference in the primary outcome of ischemic stroke, intracerebral hemorrhage, or all-cause mortality, but there was a decrease in ischemic stroke and an increase in major hemorrhage for patients who received warfarin. In a meta analysis of the 4 main RCTs, there was no difference in all-cause mortality, HFrelated hospitalization, or nonfatal MI.^[151] There was a decrease in all cause stroke and ischemic stroke and an increase in major bleeding for patients who received warfarin.^[152] The ongoing A Randomized, Double-blind, Event-driven, Multicenter Study Comparing the Efficacy and Safety of Rivaroxaban With Placebo for Reducing the Risk of Death, Myocardial Infarction or Stroke in Subjects With Heart Failure and Significant Coronary Artery Disease Following an Episode of Decompensated Heart Failure (COMMANDERHF) trial is testing the additional use of rivaroxaban vs placebo in patients with sinus rhythm, HFrEF, and a recent hospital admission (NCT01877915).

Anterior ST-elevation myocardial infarction (STEMI) and LV dysfunction has been associated with increased rates of LV thrombus and subsequent thrombotic complications. The rate of LV thrombus associated with anterior STEMI has decreased with more contemporary reperfusion strategies and DAPT. Historical rates range from 3% to 27%, depending on LV function.^{[153]–[155]} However, rates of embolization are much more difficult to quantify. There are no prospective RCTs that address the role of anticoagulation in MI with low EF. A retrospective study of 460 patients done in 2015 evaluated the role of warfarin after primary PCI for anterior STEMI.^[156] Warfarin use was at the discretion of the attending physician and 131 patients were discharged receiving warfarin; 99% were discharged receiving DAPT. The rate of death, stroke, need for transfusion, and major bleeding was higher in the warfarin group. Other cohorts have shown similar results.^[157] These data should be placed in context with emerging evidence for the use of DAPT as well as non-vitamin K antagonist oral anticoagulants in the setting of an ACS.

7.1.1.12 Anti-inflammatory medications

Several studies have shown that nonsteroidal anti-inflammatory drugs increase the risk of HF. This includes new-onset HF as well as worsening HF outcomes such as hospitalizations and even mortality. There are inconsistent data regarding the safety of individual agents in HF, however most have been associated with negative cardiovascular effects.^{[158]–[163]}

7.1.1.13 Calcium channel blockers

Most studies on the role of CCBs in HF have shown worsening in HF outcomes.^{[164]–[167]} The Prospective Randomized Amlodipine Survival Evaluation (PRAISE) and PRAISE-2 trials were RCTs that evaluated the effect of amlodipine vs placebo on all-cause mortality and/or cardiovascular hospitalization. There was no difference in either trial in terms of all-cause mortality, cardiovascular death, or hospitalizations.^[168],[169] In PRAISE, there was no overall difference between placebo and amlodipine, however, a subgroup analysis showed a reduction on cardiovascular events in patients with a nonischemic etiology of HF.^[168] In PRAISE-2, there was no significant difference between amlodipine and placebo in efficacy. The results together suggest caution when using amlodipine.

7.1.1.14 Antiarrhythmic drugs

Most anti-arrhythmic drugs (eg, amiodarone) have significant concerns related to their safety profile, especially in HFrEF, and although effective at suppressing atrial or ventricular arrhythmias, might also provoke HF decompensation and cause other adverse effects. When considering these drugs, consultation with an electrophysiologist or individual with appropriate experience and expertise in the use of these drugs is generally advisable.

7.1.2 HFpEF pharmacological treatment

Principles underpinning the pharmacological management of HFpEF include: (1) identification and treatment of underlying etiological factors implicated in the development of HFpEF; (2) identification and treatment of comorbid conditions that might exacerbate the HF syndrome;(3) control of symptoms; and (4) realization of clinically meaningful cardiovascular end points such as HF hospitalization and mortality. There remains a paucity of clinical trial data regarding specific pharmacological therapy in the HFpEF population at this time. Comorbid conditions including other chronic medical diseases are common in the HFpEF population and frequently implicated as triggers for HF decompensation, thus optimal management of these coexistent disorders, including pharmacological and nonpharmacological therapies, should be aggressively pursued.

7.1.2.1 ACEis and ARBs in HFpEF

There is, however, evidence to support the use of ARBs to reduce HF hospitalizations that draws upon secondary end point analysis from the Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity (CHARM)-Preserved trial.^[170] Among 3025 previously hospitalized NYHA class II-IV patients with an LVEF ≥ 40%, candesartan reduced the relative risk of time to first HF hospitalization by 26% compared with placebo. Moreover, a recurrent event analysis of CHARM-Preserved confirmed that this benefit extended to subsequent hospitalizations as well.^[171] Reduction in HF hospitalization has also been shown with ACEis, although the evidence is less robust and limited to data from the PEP-CHF study,^[172] which included patients 70 years of age or older with an LVEF ≥ 45%. The trial, which had a lower than anticipated event rate and high open-label crossover, did show that perindopril reduced the secondary end point of HF hospitalization by 37% at 1 year although this benefit did not persist over a mean follow-up period of 2.1 years. The I-PRESERVE trial did not show a similar benefit.^[173] The ongoing Prospective Comparison of ARNI with ARB Global Outcomes in Heart Failure With Preserved Ejection Fraction (PARAGON-HF) trial is comparing sacubitril-valsartan with valsartan on clinical outcomes in patients with HFpEF (NCT01920711).

7.1.2.2 MRAs in HFpEF

The TOPCAT trial^[174] randomized 3445 symptomatic high-risk HFpEF patients, characterized by elevations in NP levels or HF hospitalization within the previous year, to receive spironolactone (mean dose of approximately 25 mg and target dose 45 mg daily) or placebo. Patients were generally older (age older than 50 years) with relatively preserved renal function (eGFR > 30 mL/min) and serum potassium levels (K⁺ < 5.0 mmol/L). After a mean follow-up period of 3.3 years, there was no difference in the combined primary end point of cardiovascular death, aborted cardiac arrest, or HF hospitalization between groups. When considering the constituent components of the primary end point, only HF hospitalization was decreased in spironolactone treated patients (HR, 0.83; 95% CI, 0.69-0.99). Although elevated potassium levels were more prevalent in the spironolactone arm of the trial (9.1% for placebo vs 18.7% for spironolactone) this did not translate into clinical adverse events including need for dialysis or death due to hyperkalemia. Numerous prespecified and post hoc analyses of the TOPCAT trial have been performed to guide the clinical interpretation and application of these data. Notably, 28.5% of participants were enrolled in the trial on the basis of elevated NP levels. In this group, participants randomized to spironolactone had a 35% reduction in the primary end point compared with those who received placebo. This benefit of spironolactone was not observed among patients who entered the trial on the basis of a previous HF hospitalization. Marked differences in baseline demographic characteristics were observed between inclusion criteria groups; those enrolled on the basis of elevated NP levels were older, had worse renal function at baseline (higher serum creatinine and lower eGFR), and were less likely to be recruited at centres in Russia or Georgia. A significant proportion of patients recruited in the latter region might not have received the assigned study treatment and thus reliable results from TOPCAT might come mainly from the Americas.^[175] The observed geographic variation analysis showed a 15% RRR in the primary end point favouring spironolactone in patients enrolled in the Americas vs those enrolled in Russia or Georgia.^[86]

Numerous prespecified and post hoc analyses of the TOPCAT trial have been performed to guide the clinical interpretation and application of these data. Notably, 28.5% of participants were enrolled in the trial on the basis of elevated NP levels. In this group, participants randomized to spironolactone had a 35% reduction in the primary end point compared with those who received placebo. This benefit of spironolactone was not observed among patients who entered the trial on the basis of a previous HF hospitalization. Marked differences in baseline demographic characteristics were observed between inclusion criteria groups; those enrolled on the basis of elevated NP levels were older, had worse renal function at baseline (higher serum creatinine and lower eGFR), and were less likely to be recruited at centres in Russia or Georgia. A significant proportion of patients recruited in the latter region might not have received the assigned study treatment and thus reliable results from TOPCAT might come mainly from the Americas.^[175] The observed geographic variation analysis showed a 15% RRR in the primary end point favouring spironolactone in patients enrolled in the Americas vs those enrolled in Russia or Georgia.^[86]

7.1.2.3 B-Blockers in HFpEF

Although β-blockers provide a plausible physiological mechanism of action for improved outcomes by prolongation of diastolic filling time, reduction of myocardial ischemia, control of hypertension, and arrhythmia prophylaxis, the available quality of evidence and heterogeneity of findings from meta-analyses precludes a firm recommendation for use of this medication class in HFpEF at this time.^{[176]–[179]} As an example, an LVEF subgroup analysis of the Study of the Effects of Nebivolol Intervention on Outcomes and Rehospitalization in Seniors With Heart Failure (SENIORS) trial^[180] showed a 19% reduction in the combined primary end point of all-cause mortality and cardiovascular hospitalization (HR, 0.81; 95% CI, 0.63-1.04; P for subgroup interaction = 0.043) among study participants with an LVEF ≥ 35% who received nebivolol compared with placebo. However, because of the small effect size of nebivolol in the main SENIORS trial, this analysis lacks power to definitively rule out a significant interaction between outcomes of interest and EF. High dropout rates in the main trial, small sample size, and low event rate in the nonreduced EF group raise further questions about the reproducibility of these findings.

7.1.2.4 Nitrates in HFpEF

Nitrates have been broadly used in patients with established CVD, however, the role of long-acting nitrates in patients with HFpEF is unclear. The Nitrate’s Effect on Activity Tolerance in Heart Failure With Preserved Ejection Fraction (NEATHFpEF) trial^[181] enrolled 110 patients to a long-acting nitrate (isosorbide mononitrate 120 mg/d) or placebo into a 6-week crossover trial to test the efficacy and safety of this approach. There was no beneficial effect of nitrates seen in this group on biomarkers, exercise tolerance, activity level, or clinical eventsdand there was a nonsignificant trend toward a lower rate of daily activity for patients who received long-acting nitrates.

7.1.3 Implantable cardiac devices

7.1.3.1 Implantable cardioverter-defibrillator therapy

The evidence for the recommendations for implantable cardioverter-defibrillator (ICD) therapy in HF management has been discussed extensively in previous CCS HF guidelines.^{[182]–[184]} Since the publication of these updates, no new indications for ICD therapy have arisen for the general HF population; however, it is worth highlighting some of the most salient points.

7.1.3.1.1 ICD therapy in patients with HF and previous occurrence of sustained ventricular arrhythmia (secondary prevention)

Three large RCTs^{[185]–[187]} (and a subsequent meta-analysis^[188]) have compared the use of an ICD with antiarrhythmic drug therapy (primarily amiodarone) in patients with a history of life-threatening ventricular arrhythmias. Most of the patients in these trials had LVSD, and many had symptomatic HF. Although HF symptoms were not often specified as inclusion criteria in many of the trials, most patients had CAD with previous MI or nonischemic cardiomyopathy, with a mean LVEF of 30%-35%. As a primary end point, all-cause mortality was reduced in all studies in the defibrillator-treated patients compared with in the antiarrhythmic drug-treated patients (significantly lower in the Antiarrhythmics Versus Implantable Defibrillators [AVID] study^[186] and in the meta-analysis^[188]); in the secondary analyses of the studies and the meta-analysis, patients with lower EFs (< 35%), higher NYHA class (classes III or IV), and older age had a higher absolute risk of death and received greater relative and absolute benefits from ICD therapy than did patients without these risk factors. ICDs are the therapy of choice for the prevention of sudden death and all-cause mortality in patients with a history of sustained ventricular tachycardia or ventricular fibrillation, cardiac arrest, or unexplained syncope in the presence of LVSD patients with symptomatic HF, especially with LVEF < 35%, are at particularly high risk of death and stand to receive at least as much benefit as patients not meeting these clinical criteria.

7.1.3.1.2 ICD therapy in patients with HF without a history of sustained ventricular arrhythmia (primary prevention)

On the basis of the available evidence, ICD therapy for primary prevention improves survival in patients with NYHA II-III ischemic and nonischemic HF with EF < 35% and in patients with a previous MI with EF < 30% irrespective of symptom status. In contrast, ICD therapy does not provide any survival benefit early after an MI.^{[189]–[192]}

Landmark clinical trials of ICD therapy in the primary prevention setting selected patients with low LVEF; the most common LVEF cutoff was 35%, although the Multicenter Automatic Defibrillator Implantation Trial II (MADIT II)^[190] used < 30%. Although most studies did not specifically select patients with symptomatic HF, the largest study, the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT),^[189] included patients with current NYHA class II or III symptoms, and a history of HF for more than 3 months.

When considering the risk of sudden death and potential benefit from an ICD, the contribution of systolic dysfunction per se vs HF symptoms has not been fully defined. Secondary analyses of most studies have indicated that the absolute risk of sudden death, as well as the relative and absolute mortality benefits of an ICD, was greater for patients with lower LVEF (< 30%).

The contribution of HF symptoms (as distinct from LVEF) to the absolute and relative benefit of an ICD remains unclear. In MADIT II, in which patients with NYHA class I, II, or III could be enrolled, patients with greater symptoms of HF appeared to derive relatively greater benefit from an ICD.^[190] In contrast, patients in SCD-HeFT with class III HF appeared to have a smaller RRR) than those with NYHA class II symptoms.^[189]

Results on the basis of 12 RCTs (8516 patients) and 76 observational studies (96,951 patients), showed that ICD therapy was associated with a 1.2% implantation mortality and a total 3.5% annual likelihood of complications including device malfunction, lead problems, or infections. There was a 4%-20% range of annual inappropriate discharge rates.^[193] It is important to note that in RCTs that specifically selected patients early (< 40 days) after a MI, there was no significant benefit from the ICD compared with control therapy.^[191],[192]

Finally, recent evidence has called into question the benefit of ICDs for primary prevention in patients with nonischemic cardiomyopathy. Historically, the data supporting ICD use in this population have been less robust, and guideline recommendations have been largely on the basis of older systematic reviews and RCTs, including the Defibrillators in Nonischemic Cardiomyopathy Treatment Evaluation (DEFINITE)^[194] and SCD-HeFT trials, that have shown a reduction in sudden death with ICDs in patients with nonischemic cardiomyopathy. However, the recently published Danish Study to Assess the Efficacy of ICDs in Patients with Nonischemic Systolic Heart Failure on Mortality (DANISH) trial randomized 1116 patients with nonischemic HF, LVEF ≤ 35%, NYHA II-IV symptoms, and elevated NT-proBNP to ICD vs no ICD in addition to contemporary HF therapy.^[195] In this study, there was no difference between groups with respect to the primary outcome of all-cause mortality after a median follow-up of approximately 5.5 years. Importantly, ICD use was associated with a reduction in sudden cardiac death (SCD) in the overall study population, and a reduction in all-cause mortality in the subgroup of patients younger than 68 years of age. DANISH also included a very high proportion of patients treated with CRT (58%), which might have offset some of the benefits of ICD therapy. Indeed, the degree of benefit of ICDs in the setting of nonischemic cardiomyopathy is the subject of ongoing investigation; in an updated meta-analysis of primary prevention ICDs in nonischemic cardiomyopathy that included the DANISH trial a significant 23% risk reduction in all-cause mortality favouring ICD use was reported.^[196] In balance, the weight of evidence appears to favour the use of ICDs for primary prevention in nonischemic HF, however, recent data highlight the need to individualize decision-making and recommendations around ICDs, and further informs the discussion between clinicians and patients regarding the anticipated effects of this therapy.

The assessment of LVEF for ICD consideration should be performed after titration and optimization of medical therapy. It is reasonable to evaluate response to therapy and LV function at least 3 months after titration of medical therapy. In addition to cardiac status, consideration of other comorbid conditions, patient desires, and goals of therapy are essential components in the assessment for prescription of ICD therapy in this group of patients. In addition, close collaboration between the referring or HF physician and the arrhythmia specialist is essential, not only in the initial assessment of these patients, but in their follow-up. Additional considerations and related guidance is available in the CCS/Canadian Heart Rhythm Society 2016 ICD guidelines.^[197]

7.1.3.2 Device considerations in patients with HF after cardiac surgery

The rationale and evidence supporting the use of devices, ICD and CRT, in patients with HF and reduced EF have been addressed in detail in previous HF and CRT guideline updates.^[184],[198]

Although no studies to date have directly assessed the optimal timing of ICD implantation in the setting of ischemic cardiomyopathy, evidence from primary prevention trials suggests that ICDs do not confer an overall mortality benefit when implanted during, or immediately after, an acute event or revascularization.^[191] The Coronary Artery Bypass Graft (CABG) Patch trial was designed to assess whether an ICD is associated with additional survival benefit in patients at high risk for SCD who undergo coronary artery bypass graft (CABG) surgery.^[199] The negative findings in this trial were essentially mirrored in other studies of ICD after acute MI and reinforce the role of ICD therapy to be one for chronic LV dysfunction.^{[191],[192],[200]}

After revascularization, the risk of SCD continues over time,^[201] while systolic function might not improve substantially,^[202] posing a challenge in defining the optimal timing for ICD therapy.

Similarly, the optimal timing for CRT implantation in suitable candidates with ischemic cardiomyopathy has not been well defined. Key clinical trials reporting a mortality benefit with CRT excluded patients with a recent (1-6 months) MI or revascularization procedure.^{[203]–[205]} However, data from observational studies provide a rationale for considering epicardial LV lead placement at the time of CABG surgery in patients who might otherwise have an indication for CRT. Transvenous LV lead delivery via the coronary sinus is technically not feasible in approximately 10% of cases^[206]; surgical lead placement can overcome anatomical limitations imposed by the coronary sinus, with acceptable long-term lead performance and rates of clinical response similar to conventional transvenous implantation.^[207] Additionally, surgical revascularization might not have any effect on dyssynchrony, which is associated with a worse prognosis.^[208] Data from one RCT^[209] suggest that CRT using an epicardial lead implanted concomitantly with CABG is associated with improved systolic function and survival compared with CABG alone in patients with poor systolic function and evidence of preoperative device candidacy. Therefore, epicardial LV lead placement might be considered in selected patients who undergo surgical revascularization for ischemic cardiomyopathy who are likely to remain candidates for CRT after surgery.

Perioperative management of existing devices remains an important component of care; in keeping with existing guidelines, which state device deactivation is necessary before any procedure in which electrocautery, or potential for electrical interference with the device might occur.^[210] Postoperatively, re-establishment of appropriate device threshold determination and programming are recommended.

7.1.3.2.1 ICD therapy to prevent sudden death in patients with hypertrophic cardiomyopathy

Although a detailed review of specific cardiomyopathies is beyond the scope of this document, prevention of SCD in patients with an established diagnosis of HCM in particular has been an area of active study discussed elsewhere.^[211],[212]

Cardiovascular death, frequently due to sudden death, is a well-recognized complication of HCM, at approximately 1%-2% per year.^[211],[212] Well established clinical risk factors for sudden death include: previous cardiac arrest, ventricular fibrillation, or sustained ventricular tachycardia, a history of sudden death in close relatives (particularly at a young age), a history of unexplained syncope, LV wall thickness ≥ 30 mm, nonsustained ventricular tachycardia (≥ 3 beats at ≥ 120 bpm) on Holter monitoring, and blunted BP response to exercise.^[211] Although patients with multiple risk factors are at higher risk, the relative weight or importance of individual risk factors for clinical decision-making in the primary prevention setting remains the subject of ongoing study. Current guideline-based risk stratification approaches appear to have limited ability to discriminate high- vs lower-risk patients.^[213]

More recently, the HCM Risk-SCD Prediction Model^[214] is a retrospectively derived risk score that provides an absolute estimate of 5-year risk of sudden death. Attempts at validating the HCM Risk-SCD Prediction Model have yielded conflicting results in different patient populations and in different practice settings.^[215],[216] It is therefore important to recognize that current approaches to risk stratification for SCD in HCM have limitations; patient factors and other markers of risk (including specific genetic mutations, identification of LGE on CMR imaging, for example) might modify the assessment of risk in an individual patient.

There is no evidence that drug therapy reduces the risk of sudden death, even in high-risk patients. An ICD is indicated for patients with HCM who survive a cardiac arrest or have had sustained ventricular tachycardia. Although there are no prospective RCTs to guide therapy for primary prevention, there is consensus that consideration should be given to implantation of an ICD in patients with multiple high-risk factors and in patients whose estimated absolute risk (of SCD) is high.^[211],[212] Patients with a single high-risk factor should be individually assessed for ICD implantation, including a discussion of the level of risk acceptable to the individual and potential adverse effects with an ICD, such as inappropriate ICD discharges, lead complications, and infection.

7.1.3.3 CRT

Despite optimization of GDMT, LV systolic dysfunction and HF symptoms persist for many patients. Commonly, these patients have conduction delay, typically expressed as an LBBB pattern that is associated with cardiac mechanical dyssynchrony. This compromises ventricular function and is associated with poor prognosis. CRT attempts to synchronize the activation of the ventricles as well as the atrioventricular activation sequence, which leads to short-term and long-term improvements in overall LV function.

The publication of landmark trials and analyses mandated the revision of the earlier recommendations to include patients with mild HF symptoms and to place more emphasis on QRS morphology and duration, and the importance of sinus rhythm in the selection of CRT patients.^[183],[184] Further systematic reviews and long-term follow-up data from RCTs have confirmed the benefits of CRT and helped refine the selection of ideal candidates for this therapy. The updated recommendations have been harmonized with the comprehensive CCS guidelines on the use of cardiac resynchronization therapy: evidence and patient selection.^[198]

Several landmark studies have shown the effectiveness of CRT to improve morbidity and mortality in selected patients with HFrEF. Al-Majed et al. performed a systematic review of RCTs^[217] that included 25 studies of 9082 patients with LVEF ≤ 40% and compared CRT vs usual care or ICD or RV pacing alone. Pooled data from all studies showed that CRT reduced mortality by 19%. Analysis of outcomes according to NYHA functional class revealed a 17% reduction in mortality and 29% reduction in HF hospitalization among patients with NYHA I-II symptoms. Similarly, there was a 20% reduction in mortality and 35% reduction in HF hospitalization among patients with NYHA III-IV symptoms. CRT was associated with a 94.4% implantation success rate, 3.2% risk of mechanical complications, 6.2% risk of lead complications, and peri-implant mortality of 0.3%. Results of other systematic reviews, including individual patient meta-analyses^{[218]–[222]} have yielded similar findings, suggesting that CRT improves survival and HF hospitalization in a spectrum of HFrEF patients with mild or severe HF symptoms. Finally, since the publication of these reviews, the long-term follow-up of the Multicenter Automatic Defibrillator Implantation Trial With Cardiac Resynchronization Therapy (MADIT-CRT) study has been published.^[223] In this trial, 1818 patients with NYHA I and II symptoms, LVEF < 30%, and QRS ≥ 130 ms were randomized to CRT defibrillators (CRT-D) vs ICD only and median follow-up was 5.6 years. Among patients with LBBB, there was a 41% relative reduction in mortality, and among patients with right bundle branch block, a 57% relative increase in mortality. This analysis confirms the long-term mortality benefit in patients with mild HF, reduced EF, and LBBB beyond the benefits in morbidity reported in the primary trial.

An important and consistent finding in systematic reviews and in subgroup analyses of RCTs is that the benefits of CRT are greatest for patients with a broader QRS, typically defined as QRS duration > 150 ms, and for patients with a typical LBBB QRS morphology.^{[223]–[227]} It remains unclear whether patients with a relatively narrow QRS (120-150 ms) or those with non-LBBB derive any benefit from CRT, or whether other clinical factors could help select potentially appropriate candidates among these subgroups. Last, the interaction between QRS duration and morphology and its importance for CRT also warrants further evaluation; it is conceivable that patients with very broad QRS and non-LBBB morphology might derive some magnitude of benefit from CRT.^[226] Current recommendations for CRT candidate selection are therefore on the basis of the characteristics of patients included in landmark studies and on the clinical characteristics of patients shown to derive significant benefit from CRT on the basis of the totality of available data (Fig. 5).

7.1.3.3.1 CRT in patients with AF

Most RCTs that evaluated CRT included only patients in sinus rhythm, and relatively few patients with permanent AF were included in prospective randomized studies of CRT efficacy. Achieving atrioventricular synchronization is an important goal for most patients in sinus rhythm who undergo CRT, and it is therefore unclear whether patients with permanent AF who are otherwise candidates derive any meaningful benefit from CRT. To date, the Resynchronization for Ambulatory Heart Failure Trial (RAFT) is the largest RCT to include patients with AF with intact AV nodal conduction. In a substudy of RAFT^[228] the effect of CRT in patients with permanent AF was evaluated; 114 patients were randomized to CRT-D and 115 patients were randomized to ICD alone and LVEF (22.9% vs 22.3%) and QRS duration (151.0 ms vs 153.4 ms) were similar between groups. In this study, CRT was not associated with improvements in the combined end point of death or HF hospitalization or cardiovascular death, HF hospitalizations, change in 6-minute walk, or quality of life. A major limitation of this analysis is that only one-third of patients achieved biventricular pacing 95%, and only 1 patient underwent AV nodal ablation to effect 100% biventricular pacing.

Indeed, observational data strongly suggest that outcomes in AF patients who receive CRT are associated with the degree of biventricular pacing achieved, and that differential effects in survival might be seen when biventricular pacing achieved is < 98% vs > 98%.^[229] A meta-analysis of observational studies of AV node ablation vs pharmacologic rate control in AF and CRT (1256 patients; 644 with AV node ablation, 798 without AV node ablation) suggested that AV nodal ablation is associated with a higher degree of biventricular pacing (100% vs 82%-96%), reduced mortality, and lower rates of CRT nonresponse compared with pharmacologic rate control.^[230] Ongoing prospective RCTs, including the multicentre Resynchronization/Defibrillation for Ambulatory Heart Failure Trial in Patients With Permanent AF (RAFT-PermAF) should help refine the role of CRT in patients with AF who would otherwise be suitable candidates.

7.1.3.3.2 CRT in patients with RV pacing and reduced EF

The use of CRT in patients with LVSD who require permanent ventricular pacing, or in patients with suspected RV pacing-induced HF has been investigated.^{[231]–[233]} Although the prognostic significance of RV pacing-induced dyssynchrony vs intrinsic LBBB-related dyssynchrony is uncertain, a subgroup of patients with frequent RV pacing will experience worsening of LV function, particularly in the setting of abnormal LVEF and HF at baseline. The results of these studies suggested that CRT in this clinical situation improves LV function, symptoms, and exercise capacity.^[231] To address this issue further, the Biventricular Versus RV Pacing in Patients with LVSD and Atrioventricular Block (BLOCK HF) study randomized 691 patients with LVEF ≤ 50% and heart block to CRT vs RV pacing (with an ICD or pacemaker as indicated).^[234] Patients in this study had a mean LVEF of 40%, and > 80% had NYHA class II or III symptoms. After a mean follow-up of 37 months, CRT was associated with fewer primary outcome events including the composite of death, urgent care visit for intravenous (I.V.) HF therapy, or an increase in LV end systolic volume index ≥ 15% (HR, 0.74; 95% CI, 0.60-0.90). The benefits observed with CRT were driven by reductions in HF events. Notably, pacing percentage in both study groups was > 97% and serious adverse events occurred in 14% of patients, mainly related to lead complications. Overall, it appears that patients similar to those included in the BLOCK HF study derive significant benefits with CRT compared with RV only pacing with respect to HF events, but the potential for procedural complications needs to be considered carefully for individual patients.

7.1.3.3.3 CRT in patients with narrow QRS

Compared with ICD alone, CRT has not been associated with improvements in mortality or HF hospitalization, and there is a suggestion of increased harm with CRT in some studies.^{[235]–[239]}

7.1.4 Advanced HF management strategies

Although the term, advanced HF has many definitions, to guide clinicians as to which patients should be considered for advanced HF management (such as but not limited to cardiac transplantation, MCS, or palliative care) the following is a general guide. Cardiac transplantation is well established in Canada and further guidance is available at https://ccs.ca/en/cctn-home. Cardiac transplantation assessment is typically done by a multispecialty, multidisciplinary team in a specialized setting, using Canadian and international guidance for appropriate workup and eligibility.

Patients with advanced HF to be considered for advanced HF management strategies include those who, despite optimal treatment, continue to exhibit progressive/persistent NYHA III or IV HF symptoms and accompanied by more than one of the following:

• LVEF < 25% and, if measured, peak exercise oxygen consumption < 14 mL/kg/min (or less than 50% predicted).
• Evidence of progressive end organ dysfunction due to reduced perfusion and not to inadequate ventricular filling pressures.
• Recurrent HF hospitalizations (≥ 2 in 12 months) not due to a clearly reversible cause.
• Need to progressively reduce or eliminate evidencebased HF therapies such as ACEis, MRAs, or β-blockers, because of circulatory-renal limitations such as renal insufficiency or symptomatic hypotension.
• Diuretic refractoriness associated with worsening renal function.
• Requirement for inotropic support for symptomatic relief or to maintain end organ function.
• Worsening right HF (RHF) and secondary pulmonary hypertension.
• Six-minute walk distance < 300 m. Increased 1-year mortality
• (eg, > 20%-25%) predicted by HF risk scores
• Progressive renal or hepatic end organ dysfunction.
• Persistent hyponatremia (serum sodium < 134 mEq/L).
• Cardiac cachexia.
• Inability to perform activities of daily living.

It should be noted that most patients will have a number of the listed criteria and there is no single criterion that determines candidacy for cardiac transplantation, MCS, or palliative care. Patient preferences should be incorporated into the decision process when assessing further choices.

7.1.5 MCS

7.1.5.1 What is mechanical circulatory support?

MCS is a group of technologies that increase forward cardiac output in patients.^[240] MCS therapies consist of ventricular assist devices that augment or replace the ventricle. They may be used to assist the right ventricle, left ventricle, or both ventricles.^[241] The choice depends on the clinical presentation, and can be divided into 2 categories-temporary circulatory support and long-term devices. Details of the purpose of MCS, the decision process, description of patient profiles, management, and other issues are outlined in sections 7.1.5.2-7.1.5.7 of the Supplementary Material, and in Tables 15-17.

7.1.6 Exercise and rehabilitation

Exercise intolerance is recognized as a hallmark of HF. It is now understood that exercise intolerance in HF has a multifactorial etiology and that parameters such as intracardiac filling pressures and LVEF might not be reliable predictors of exercise capacity. Changes in the periphery and LV function are both important determinants of exercise capacity. Therefore, it is rational that exercise training could potentially benefit patients with HF.

There have been several systematic reviews and metaanalyses that show the benefits of exercise training for patients with HF.^{[243]–[245]} There has been one large RCT that has shown the benefits of exercise training.^[246] In the Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) study 2331 medically stable patients with HF were randomized to regular exercise training or usual care. There was a nonsignificant 7% relative reduction in the primary outcome of all-cause mortality or hospitalization. After adjusting for the key covariates of duration of the cardiopulmonary exercise test, LVEF, Beck Depression Inventory II score, and history of AF or flutter along with HF etiology exercise training was found to produce an 11% relative reduction of all-cause mortality or all-cause hospitalization. There was no difference in adverse events between the 2 groups. A systematic review examined the effectiveness of exercise-based rehabilitation in 33 trials with 4740 patients with predominately HFrEF.^[243] There was no reduction in allcause mortality with up to 1 year of follow-up, and a trend to reduction in all-cause mortality in trials with more than 1 year of follow-up (6 trials; 2845 participants; relative risk, 0.88; 95% CI, 0.75-1.02). Exercise training did reduce all-cause hospitalization and HF hospitalizations. Importantly, no safety concerns were raised in any of the studies.^{[245]–[248]}

Exercise training results in increases of exercise capacity.^[249],[250] A meta-analysis of studies that directly measured peak volume of oxygen (VO₂) reported an average 17% improvement in peak VO₂.^[245] The evidence for quality of life improvements in patients with HF exposed to exercise training is further supported by multiple studies.^[245],[251] Although there have been a great variety of types of exercise training strategies most have involved moderate to vigorous exercise as was prescribed in the HF-ACTION study.^[250]

Although the dose of physical activity that conveys cardiovascular and other health benefits is difficult to categorically quantify, there is support for as little as 15 minutes per day of moderate-intensity physical activity with further dose response thereafter.^[252] Piepoli et al.^[253] have provided an overview about the practical approaches to exercise training for patients with HF.

The role of exercise training in HFpEF patients is less well established. However, the available data suggest exercise training has benefits that include improvements in exercise capacity and quality of life.^{[244],[254]–[256]} Studies have also shown that exercise training can take place in patients with ICD or CRT therapy. These studies have shown that properly prescribed and monitored exercise training can safely result in improvements in exercise capacity in patients with an ICD and/or CRT therapy (Table 18).^[257],[258]

7.1.7 Important nonpharmacological and nondevice management options

The basic treatment of HF has included advice on dietary salt and fluid restriction. The evidence to support these concepts is scarce and some evidence suggests the opposite of current clinical practice.^{[259]–[262]}

Dietary sodium consumption for patients with HF remains controversial. Several cohort studies and 1 RCT suggest that lower dietary sodium intake is associated with better clinical outcomes.^{[260],[262],[263]} Other studies suggest that the combination of dietary sodium restriction and high-dose diuretics with or without saline infusions can be deleterious, summarized by Gupta et al.^[264] An ongoing RCT will provide guidance on this topic (NCT02012179).

Severe fluid restriction is not only difficult to maintain but could also have deleterious effects without additional benefit. Three hospital-based RCTs in this area suggest no additional benefit of the combination of fluid restriction with or without sodium restriction.^{[265]–[267]} Special consideration for severe water restriction should be considered for hyponatremic patients with hypervolemia and applied sparingly. High-quality data are lacking on this topic, and no high-quality evidence exists in the ambulatory care environment. Thus, for patients admitted to hospital or as outpatients, allowing liberal fluid intake is reasonable.

Alcohol consumption should be limited for all patients with HF, and if it is believed to be responsible and/or contributing to the syndrome it should be avoided altogether because there is a dose-dependent effect and individual susceptibility to the deleterious effects of alcohol.^[268]

Because smoking has been linked to the progression of CAD all attempts should be done to promote smoking cessation, even if HF is not present. Nicotine replacement therapy and/or other smoking cessation therapies are acceptable for most patients with HF. There is limited evidence of the effects of e-cigarettes (“vaping”) or medical marijuana for patients with established HF.

7.2 Cardiovascular comorbidities

7.2.1 Atrial fibrillation

AF and HF share common risk factors and frequently coexist.^[270],[271] Up to 50% of patients with HF might develop AF; the reported prevalence rates of AF among HF cohorts varies significantly and is largely dependent on the clinical setting (acute vs community), the extent of LV dysfunction, NYHA functional class, and the use of background HF therapies.^[272]

The presence of AF is associated with a worse prognosis in terms of overall survival^[273],[274] as well as risk of stroke.^[275] AF might also exacerbate the HF syndrome through a number of mechanisms including: (1) decreased cardiac output secondary to loss of atrial systole; (2) increased myocardial oxygen consumption and decreased coronary perfusion during periods of rapid ventricular response; (3) neurohormonal activation; and (4) the development of tachycardia-induced cardiomyopathy.^[276] Further details pertaining to primary prevention, rate and rhythm control, and anticoagulation can be found in sections 7.2.1.1-7.2.1.3 of the Supplementary Material.

7.2.2 CAD and revascularization

Nearly 60% of patients with chronic HF suffer from CAD, and approximately 15% of AHF cases occur in the setting of an ACS.^[278],[279] Despite the coexistence of CAD and HF, few clinical trials have been performed that can inform optimal care for patients with these conditions. Several areas exist in which high-grade evidence is lacking, such as coronary revascularization in the setting of HFpEF. Although ample evidence exists to support PCI in patients with HF due to ACS,^[280] there is limited evidence to support its use in the setting of chronic HF to reduce adverse clinical outcomes.^[281] A detailed discussion of different imaging modalities, an approach to the diagnosis of CAD, and the perioperative management and the approach to revascularization are covered in sections 7.2.2.1-7.2.2.5 of the Supplementary Material, and Figures 6 and 7.

7.2.2.4 Disease management, referral, and perioperative

7.2.2.5 Surgical revascularization for patients with CAD and HF

The approach to the decision of coronary revascularization in patients with HF is illustrated in Figure 7. CABG surgery is indicated in adult patients with symptoms of angina, a history of HF in association with LV dysfunction (LVEF < 35%), graftable coronary arteries, and who have an otherwise good life expectancy. This recommendation is on the basis of historical data from earlier landmark clinical trials comparing medical and surgical therapy, which identified a survival benefit with CABG in patients with triple vessel CAD along with ventricular dysfunction.^{[283]–[286]} Further details of the trials in this area are available in section 7.2.2.5 of the Supplementary Material.

The Surgical Treatment for Ischemic Heart Failure (STICH) trial sought to address 2 hypotheses: (1) does CABG improve survival in combination with optimal medical therapy for patients with HF and CAD (LVEF < 35%) who are acceptable candidates for cardiac surgery; and (2) does the additional use of SVR of an akinetic/dyskinetic anterior wall provide better outcomes than isolated CABG for eligible individuals.^[287] This study evaluated patients with ischemic cardiomyopathy with or without HF symptoms and randomized them into 3 groups, namely, optimal medical therapy and CABG alone, CABG with the SVR procedure, or neither procedure. Details pertaining to the 2 hypotheses are available in the Supplementary Material.

A major concern regarding surgical revascularization in patients with LV dysfunction is a greater rate of operative mortality. A meta-analysis of 26 observational studies (3621 patients) with a preoperative LVEF < 35% showed an operative mortality of 5.4%.^[288] The 2 risk calculators for surgical mortality have been recently updated: EuroSCORE II (http://www.euroscore.org/calc.html) and the (Society of Thoracic Surgeons) STS score (http://riskcalc.sts.org/ STSWebRiskCalc273/de.aspx).

The Canadian Association of Cardiac Rehabilitation and CCS joint position statement includes routine cardiac rehabilitation for patients with HF who successfully complete CABG surgery.^[282]

7.2.3 Right heart failure

RHF is defined as the clinical syndrome in which the right ventricle function is impaired secondary to any structural or functional cardiac disorders leading to inadequate blood flow through the pulmonary circulation at a normal central venous pressure.^[289] The most common reason for RHF is left-sided HF but occasionally RHF might occur as pure right-sided HF (Table 19).

In HF, RV dysfunction is a strong predictor of mortality.^[290],[291] In a study of 377 patients with chronic HF who underwent right heart catheterization, 75% of patients had reduced RV function which was an independent risk of death.^[292] In another study of 250 consecutive patients with dilated cardiomyopathy it was shown that reduced RV EF < 45% using magnetic resonance imaging (MRI) was an independent predictor of transplantation-free survival and worse prognosis.^[293] The underlying pathophysiology of RHF might include venous congestion, RV enlargement, increased pulmonary artery pressure (PAP) and tricuspid or pulmonary valve abnormalities.

The clinical presentation of RHF is variable but typically involves fluid retention (ascites, peripheral edema), decreased systolic reserve or low cardiac output (fatigue, exercise intolerance), atrial or ventricular arrhythmias, and hypotension. Gastrointestinal symptoms like anorexia, bloating, nausea, and constipation are very common in patients with advanced RV failure. There are several medical conditions that mimic or coexist with RHF including liver cirrhosis, nephrotic syndrome, and renal failure with volume overload.

Of all physical signs of RHF, an abnormal jugular venous pressure is almost always present. In more advanced cases pitting edema, ascites, and liver enlargement are present. In the absence of elevated venous pressure, peripheral edema and ascites are unlikely to be due to RHF.

All patients with suspected RHF should undergo transthoracic echocardiography. CMR imaging has become the test of choice for noninvasive assessment of RV size, function, viability, potential etiology, and mass.^[294] In selected patients with RHF, right heart catheterization should be considered to help determine etiology of RHF and provide PAP, PCWP, and PVR. PH is defined by a mean PAP of ≥ 25 mm Hg and increased PVR of > 3 Wood units with a normal PCWP < 15 mm Hg.^[295]

There are very few RCTs that addressed the management of isolated RHF, recognizing that the most common cause of RHF is left heart disease. Generally, diuretics are the mainstay of therapy. Because patients with RHF might have normal or even low LV filling pressures, cautious use of diuretics is the key, because excessive diuresis can result in prerenal azotemia, hypotension, and exacerbation of arrhythmias. As such, it is not uncommon to see combination diuretic therapy to avoid excessive potassium loss or alkalosis.

Studies designed to see if treatment for LHF also ameliorates RHF failed to show benefit because these studies were underpowered or mechanisms of injury and repair might differ.^{[296]–[299]} Patients with RHF secondary to congenital heart disease or secondary to PH should be referred early to specialized clinics for investigations and management.^[300],[301]

Cor pulmonale term is used in cases of RHF associated with pulmonary hypertension as a result of lung disease. Determination of the etiological factor is of utmost importance because several therapies specific to the underlying cause have been developed (Table 20). For these reasons, patients with HF and PH (without LV failure) should be referred to centres with experience and expertise in the management of this disorder. In particular, patients with congenital heart disease might present with RHF due to a wide variety of specific anomalies or surgical residua. When identified, these patients should be referred to an adult congenital heart disease centre.^[301]

The diagnosis of cor pulmonale should be considered in all patients with lung disease and symptoms and/or signs of RHF. The tests used for the diagnosis of cor pulmonale include chest x-ray, ECG, echocardiogram (ECHO), CT scan, ventilation/perfusion lung scanning, MRI, pulmonary function test, and right heart catheterization. In some cases lung biopsy might be required to determine the underlying cause for cor pulmonale. The treatment of PH and cor pulmonale is determined by the etiology and specific management of these patients is addressed in disease-specific guidelines.^[302],[303] Patients with PH and cor pulmonale should be referred to the centres with appropriate expertise for the confirmation of diagnosis, vasoreactivity testing, and institution of appropriate treatment.

7.2.3.1 Arrhythmogenic right ventricular cardiomyopathy

ARVC is a genetic form of cardiomyopathy characterized by fatty/fibrofatty infiltration of the myocardium affecting mostly the right ventricle but occasionally the left ventricle too.^[304] ARVC is an autosomal dominant inherited disease with variable penetrance and expression. Prevalence of ARVC is estimated to be 1:1000-1:5000^[305],[306] affecting men more frequently than women with a ratio of 1:3.^[307]

ARVC should be suspected in a patient with unexplained RV dysfunction, dilation, or RHF, a history of ventricular tachyarrhythmia (particularly of LBBB morphology) or syncope, characteristic ECG changes (eg, epsilon waves), a family history suggestive of syncope or sudden death, and in young people or athletes with a history of syncope or cardiac arrest during exercise or sports activities.

In 2010, the 1994 European Society of Cardiology/International Society and Federation of Cardiology joint task force criteria for ARVC were revised by introducing quantitative measures into the criteria.^[308] Current task force criteria for ARVC diagnosis include diagnostic criteria on the basis of 6 categories: typical ECG findings (eg, epsilon waves), ventricular arrhythmias (left bundle block morphology), morphological and functional changes in the right ventricle, histopathology, family history, and genetic findings. On the basis of these criteria, 3 levels for the ARVC diagnoses were established. Major and minor criteria for ARVC are listed in Table 21.^[308]

Echocardiography and CMR imaging are the most commonly used tests for the diagnosis of ARVC.^[309] Strain echocardiography is one of the newest modalities and it appears to be very effective in the assessment of RV function.^[310] CMR imaging if available is a preferred test to echocardiography for ARVC.^[311] It can provide very accurate information on LV and RV function and the presence of intramyocardial fat and fibrosis via LGE. In some patients myocardial biopsy might be useful but diagnostic sensitivity is very low because of the patchy distribution of the disease.

Genetic testing can be useful in patients with a diagnosis of ARVC however, an existing known gene mutation is present in only approximately 50% of cases. Therefore, negative gene testing does not rule out ARVC.

The primary goal of treatment of ARVC is to identify high-risk patients to reduce the risk of sudden arrhythmic death. The secondary goal of treatment is to manage symptoms of ventricular arrhythmias and RHF. To date, there have been no RCTs to determine the efficacy of pharmacological and device therapies on the prevention of sudden death. Patients with ARVC and RV or LV failure should be treated with standard medical therapy. Antiarrhythmic medications have been used frequently to decrease frequency of ICD discharges.^[312] β-blockers have shown effectiveness in reducing adrenergically stimulated arrhythmias.^[313]

ARVC patients at high risk for sudden death on the basis of a clinical profile with one or more risk factors for sudden death should be considered for ICD as a primary prevention.^[314] For secondary prevention of sudden death in the ARVC population, ICDs are indicated.^[315] Cardiac transplantation is an option for eligible patients with advanced ARVC and intractable HF or ventricular tachyarrhythmia.

Exercise has been linked to the increased risk of ventricular tachycardia secondary to the elevated levels of catecholamines during physical activity in the setting of ARVC.^[316] Patients with suspected or confirmed ARVC should avoid vigorous physical activity like competitive sports, or regular activities associated with symptoms.^[317],[318]

All patients with ARVC should be referred to experienced centres with electrophysiology services and genetic counselling.

7.2.3.2 Constrictive pericarditis

Constrictive pericarditis is an uncommon disease of pericardium resulting from chronic inflammation and fibrosis leading to impaired diastolic filling of the ventricles with or without reduced systolic function. In developing countries, the most common reason for the constrictive pericarditis is tuberculosis.^[319]

Constrictive pericarditis should be suspected in patients with unexplained RHF in whom there is a history of pericardial disease or predisposing pericardial injury. The most commonly reported symptoms are peripheral edema and exertional dyspnea. Almost all patients have abnormal jugular venous pressure. It is elevated with very prominent, deep y-descent and frequently there is an increase in the jugular venous pressure with inspiration (Kussmaul sign). Pulsus paradoxus is frequently present as are an enlarged liver and ascites.

The diagnosis of constrictive pericarditis is frequently delayed because clinical presentation is often atypical and might mimic other causes of the RHF.^[320] Chest x-ray is very helpful in diagnosis of constrictive pericarditis because in 27% of cases, calcification can be seen in the pericardium.^[321] In cases in which diagnosis of constrictive pericarditis is suspected, a special “constriction protocol” should be requested when ordering an ECHO.^[322] It should include a focus on the motion of the ventricular septum, variation in the mitral inflow velocity, variation in the hepatic vein profile, and tissue Doppler assessment of mitral annular velocities with simultaneous recording of respiration and tissue strain.^{[323]–[325]} A CT scan is helpful in the assessment of pericardial thickness and calcifications. CMR imaging can provide anatomical details, hemodynamic information, and an assessment of pericardial inflammation. When noninvasive evaluation is indeterminate, cardiac catheterization with hemodynamic assessment is the test of choice.^[326]

Management includes treatment to relieve symptoms of RHF, control secondary arrhythmia, and provide timely surgical consultation for pericardiectomy. Transient constrictive pericarditis post cardiac surgery can resolve spontaneously or after a short course of anti-inflammatory medications or corticosteroids.^[327]

All patients with constrictive pericarditis should be referred to experienced centres with advanced cardiac imaging, catheterization, and surgical availability for assessment and treatment. All symptomatic patients with chronic constrictive pericarditis should be considered for pericardiotomy. Current surgical mortality rates average 6%-12%,^[328],[329] but can be elevated further if there is coexisting myocardial damage, extensive pericardial calcification (‘outer porcelain heart’), or previous mediastinal radiation.

7.3 Noncardiovascular comorbidities

Although treatments that improve survival, exercise capacity, and reduce hospitalizations have been established for HFrEF, the increased complexity of patients and their comorbidities often confound treatment. The latter issues have proven even more challenging for the HFpEF population. Comorbid conditions become risk factors for future deterioration and might contribute to clinical deterioration, complicate management, and are often associated with poor prognosis.

7.3.1 Anemia and iron deficiency

Anemia is often defined according to knowledge of normal, age- and sex-specific values of Hb, or hematocrit. The World Health Organization (WHO) defines anemia as a Hb level < 130 g/dL for men and 120 g/dL for women^[330]; other definitions also exist.

Anemia has a prevalence ranging from 10% to 68%^[331],[332]; a wide range attributable to the various definitions used and populations studied. Factors associated with anemia in chronic HF include older age, diabetes, chronic kidney disease (CKD), more advanced HF, recent HF hospitalizations, signs of HF, higher levels of neurohormones and inflammatory markers, exercise intolerance, and reduced quality of life.^{[333]–[337]}

The prevalence of anemia is similar whether EF is reduced or preserved.^[334],[338] Although anemia in HF was once thought to be almost solely attributable to CKD; anemia and CKD are now both established independent predictors of mortality and hospitalizations for HF.^[339],[340] Even small reductions in Hb levels are associated with worse outcomes and even mild anemia is associated with worsening of symptoms, increased NYHA class, and impairment in functional capacity and quality of life. Anemia is associated with higher costs of hospitalization for patients with HF.^[341],[342] A decline in Hb over time is also associated with mortality and morbidity.^[335] Therefore, there has been continued interest for more than a decade in finding appropriate treatments for anemia in patients with HF, although the underlying pathophysiology is complex and remains only partly understood. The main mechanisms include CKD, inflammation, hemodilution, absolute or relative iron deficiency (ID), rarer nutritional deficiencies (vitamin B12, folic acid, thiamine), gastrointestinal blood loss, and a few therapeutic agents with a generally low effect on Hb levels (eg, ACEis) (Table 22). In HFrEF, increased myocardial remodelling, inflammation, and volume overload have been described as the hallmarks of patients with anemia and HF.^[333]

In the CHARM studies, the effect of anemia on the primary outcome of cardiovascular death or HF hospitalization was slightly less in HFpEF than in HFrEF.^[334] However, observational and population-based studies suggest that the effect of anemia on prognosis appears similar for HFpEF and HFrEF.^[343],[344]

As for any other group of patients, reversible causes of anemia should be sought and treated. Beyond this first step, treatment options for patients with HF include evaluation of the contribution of volume overload, of concomitant medications (eg, antiplatelet agents, anticoagulants, and especially their combination); identification of ID and treatment with oral or intravenous (I.V.) iron supplements; and optimization of HF therapy.

7.3.1.1 Iron deficiency

Iron is necessary for optimal hematopoiesis but also plays a central role in oxygen transport (Hb), storage (myoglobin), cardiac and skeletal muscle metabolism; synthesis, and degradation of proteins, lipids, ribonucleic acids, and for mitochondrial function. ID, either absolute or functional, has emerged as another independent predictor of outcomes^[332],[345] and a major contributor to exercise intolerance in HF, even in the absence of anemia.^[346] ID might be detected before anemia appears, thus providing an earlier opportunity in view of improving outcomes. In a cohort of 1506 patients with chronic HF, ID, defined as a ferritin level < 100 mg/L or ferritin 100-299 mg/L if the transferrin saturation was < 20%, had a prevalence of 50%.^[345] It is estimated that 60% of patients with HF with anemia and 40% of those without anemia have ID.^[347] This definition might underestimate the prevalence of ID.

Functional ID is seen when there is a deficit in the mobilization of iron from tissues while iron stores are normal, which is frequent in chronic diseases with inflammation.^[348],[349] Hepcidin, soluble transferrin receptor and reticulocyte Hb have been proposed as more sensitive indices to evaluate ID.^{[347],[350],[351]}

A meta-analysis by Avni and colleagues^[352] had reported improvements in quality of life, 6-minute walk distance, and all-cause hospitalization with iron replacement therapy. Qian and colleagues^[353] reported on the effects of I.V. iron therapy on clinical outcomes in HF, in a second meta-analysis, including a total of 907 patients from 5 clinical trials. There were no increases in adverse events with I.V. iron therapy, using iron sucrose or ferric carboxymaltose (FCM) in these studies. Most patients included in that meta-analysis came from 2 larger trials, Ferric Carboxymaltose Evaluation on Performance in Patients With Iron Deficiency In Combination With Chronic Heart Failure (CONFIRM-HF) (n = 301)^[354] and Ferric Carboxymaltose Assessment in Patients With Iron Deficiency and Chronic Heart Failure (FAIR HF) (n = 459).^[355] The CONFIRM-HF trial^[354] included 301 patients (251 completed the trial) with moderate HF symptoms (NYHA class II-III), LVEF ≤ 45%, elevated BNP or NT-proBNP, and ID. I.V. iron was given as a FCM solution equivalent to 500 or 1000 mg of iron. At week 24, the primary end point of change in the 6-minute walk distance improved more in the FCM group (difference, 33 ± 11 m; P = 0.002). The benefit was maintained up to 52 weeks. Fatigue, NYHA class, and quality of life scores also improved on through week 52, and FCM was associated with a reduction in the risk of hospitalization due to worsening HF. The Effect of Ferric Carboxymaltose on Exercise Capacity in Patients With Iron Deficiency and Chronic Heart Failure (EFFECT-HF) study (n = 173) also recently reported benefits of I.V. FCM on peak VO2 compared with standard of care in patients with HF, EF ≤ 45%, and ID.^{[353],[356],[357]}

In patients with HF with anemia and ID in whom iron repletion is being contemplated, iron supplementation should be considered to improve functional capacity.^[358] Oral iron has no known efficacy in this context, and therefore I.V. iron should be given considering the recently presented results from the Iron Repletion Effects on Oxygen Uptake in Heart Failure (IRONOUT HF) study. In that study (n = 225), high-dose oral iron minimally repleted iron stores and did not improve peak VO2 in patients with ID and HFrEF.^[359] It is essential that the causes of absolute ID (such as iron loss) have also been investigated and treated by other means when possible (eg, endoscopy for gastric ulcer, colon cancer, etc). In such cases, concomitant I.V. iron therapy can reduce the time needed to correct anemia, as well as the need for transfusions.

Further evidence is warranted regarding the effect of I.V. iron repletion on major cardiovascular events (namely death), especially for nonanemic patients with HF and those with HFpEF. Cost-effectiveness also requires further validation with various I.V. iron formulations.

7.3.1.2 Erythropoiesis-stimulating agents

ESAs have been studied as a potentially promising class of agents to increase Hb in HF, considering not only CKD but multiple proposed pleiotropic properties of such agents. The 2 largest trials on ESAs in HF were the Study of Anemia in Heart Failure Trial (STAMINA-HeFT)^[360] and the Reduction of Events with Darbepoetin Alfa in Heart Failure (RED-HF) trial.^[361] These 2 trials, and a meta-analysis^[362] failed to show benefits on mortality, cardiovascular events, and hospitalizations. In RED-HF, a significant increase in thromboembolic events was reported in patients with Hb levels > 130 g/dL.^[361] On the basis of the results of those studies, it is unlikely that another morbidity or mortality study will be undertaken with results that will support the use of ESAs in HF.

Although ESA administration is common practice in advanced CKD, the debate continues on target Hb and the effect on quality of life, which is likely higher for those with lower Hb levels.^[363] Patients with advanced CKD (eGFR < 30 mL/min/1.73 m²) should be managed in concert with nephrologists, especially when hemodialysis is contemplated. In those cases, as well as in some hemato-oncologic conditions, ESA therapy might be an option.

7.3.2 Diabetes(treatment)

7.3.2.1 Glycemic control in patients with diabetes and HF

With the available evidence, an intensive glycemic control strategy cannot be recommended for all patients with diabetes. Instead, each individual should be assessed for his or her optimal glycemic target for the prevention of macrovascular events.

7.3.2.2 Pharmacological therapy for type 2 diabetes in patients with HF

Metformin. Metformin is still considered first-line pharmacological therapy for type 2 diabetes.

SGLT-2 inhibitors. Ongoing trials of SGLT-2 inhibitors will inform the use of the class of agents in patients with established HF. Few patients in the EMPA-REG OUTCOME trial had HF at baseline (approximately 10%), however, patients with HF had results similar to those in the overall trial.^[27] There are ongoing trials of SGLT-2 inhibitors specifically enrolling patients with HF that might inform future recommendations.

DPP-4 inhibitors. There are no high-quality HF-specific studies from which to provide guidance for patients with established HF.

GLP-1 agonists. Several small trials have tested the additional use of a GLP-1 agonist in patients with HF. None have shown any additional benefit in patients with HF, with or without diabetes.^[364],[365]

Thiazolidinediones. Two such thiazolidinedione drugs (pioglitazone and rosiglitazone) have each been shown to increase the risk of HF events and should be avoided in patients with HF, as summarized in section 4.3.1. Glycemic Control in Diabetes to Prevent HF.

7.3.3 Cardiorenal syndrome

Cardiac and renal dysfunction often occurs in concert with hemodynamic, neurohormonal, vascular, and hematologic consequences. Previously, renal dysfunction was thought to represent merely comorbidity in patients with advanced HF. It is increasingly recognized that cardiac and renal interaction is complex. Cardiorenal syndrome (CRS) refers to interactions in which renal dysfunction and HF interact and mutually reinforce each other.^[370] Mechanistic hypotheses are discussed elsewhere.^[370],[371] Elevated intra-abdominal pressure as well as central venous pressure are linked to rising serum creatinine levels. Elevated renal vein hydrostatic pressure is an important mechanism in volume-expanded patients.^[372],[373]

When managing these patients the Acute Dialysis Quality Initiative (ADQI) definition of CRS should be followed (Table 23).^[371],[374]

A meta-analysis of observational studies confirms that patients with HF with moderate to severe renal dysfunction have more than a twofold increase in relative mortality risk.^[366] The presence of HF in a hemodialysis population portends a poor prognosis with mean survival of < 36 months.^[367]

Creatinine clearance calculated using Cockcroft and Gault,^[368] Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI), or the Modification of Diet in Renal Disease formula^[369] is used to estimate the glomerular filtration rate (GFR). Standardized and validated criteria might be useful to estimate acute changes in renal function at the bedside for patients with AHF when renal injury is a possibility. The National Kidney Foundation Kidney Disease Outcomes Quality Initiative guidelines should be used for the classification, and evaluation of CKD (Table 24).^[369] It is important to recognize that markers used in the assessment of HF including BNP and NT-proBNP might need to be interpreted with caution in the presence of acute renal failure or end-stage renal disease.

7.3.3.1 Role of hemodialysis

Hemodynamic stress, metabolic changes, and electrolyte shifts often occur in patients receiving hemodialysis and might be poorly tolerated. Dialysis should be considered in patients with HF with signs and symptoms of complications of renal failure.^[375] Early consultation with a nephrologist is recommended for patients with acute kidney injury and in situations in which dialysis is being considered.

Before initiating dialysis, clinicians should be aware of the poor prognosis of patients with HF and end-stage renal disease (see section 8. Community Management of HF). When initiated, clinicians and patients might have difficulty accepting the need to discontinue or continue dialysis. Common complications associated with dialysis include dehydration and electrolyte imbalance, which might lead to angina, hypotension, or arrhythmias if left untreated. A reduction in medical therapy might be necessary for effective hemodialysis to occur, and there should be caution when reintroducing these at a later time.

Medications specific to HF should be continued for patients treated with hemodialysis when possible. An RCT of carvedilol in 114 hemodialysis patients with a low LVEF and HF symptoms showed a significant reduction in mortality or hospitalization over 2 years and improvement in NYHA and LV remodelling.^[376] Cohort data suggest that ACEis are associated with a reduction in all-cause mortality.^[377] A single RCT of 397 high-risk patients without HF receiving hemodialysis showed a trend toward a lower event rate with the use of fosinopril.^[378] Similarly, ARBs have been tested in an openlabel trial in hemodialysis patients^[379] and were not reported to reduce clinical events. One randomized trial of 332 patients with HF receiving an ACEi and hemodialysis showed a significant all-cause mortality reduction over 3 years with the ARB telmisartan but with significantly more hypotension and dropouts in the treatment arm.^[380] These results do not alter previous recommendations on combination therapy with an ACEi and an ARB. Aldosterone blockade has been evaluated in 3 small cohorts for safety^{[381]–[383]} and 1 small RCT of 16 patients with HF without significant benefits.^[384] There are limited safety data and no efficacy data favouring digoxin for patients with HF receiving hemodialysis.

7.3.3.2 Role of renal transplantation

Renal transplantation is an option for selected candidates with HF according to internationally accepted guidelines.^[385] Three cohort studies have highlighted the importance of the cardiorenal interaction in patients who have undergone renal transplantation and subsequently had improvement in symptoms, LV function, and remodelling.^[386],[387] Postoperative adverse cardiac events were low (< 5%) in these patient cohorts.

7.3.3.3 Role of ultrafiltration

Diuretic therapy is the mainstay for relief of volume overload for AHF.^[388] However, evidence-based data are sparse and diuretics have adverse effects such as activation of the neurohormonal cascade, electrolyte depletion, and renal injury.^[389],[390] Venovenous UF has been evaluated as an alternative therapy in this setting. Potential advantages of UF include greater control over the rate and volume of fluid removal, greater net loss of sodium, and less neurohormonal activation.^[391]

In the multicentre randomized controlled Relief for Acutely Fluid-Overloaded Patients With Decompensated Congestive Heart Failure (RAPID-CHF) trial the feasibility, safety, and efficacy of early UF vs usual care in the management of AHF was assessed.^[392] Early UF resulted in a trend toward greater weight loss and fluid removal at 24 hours. UF was well tolerated and the median volume of ultrafiltrate removed during a single 8-hour course of UF was 3213 mL. Dyspnea and HF symptoms were improved in the UF group at 48 hours.

In the Ultrafiltration versus Intravenous Diuretics for Patients Hospitalized for Acute Decompensated Congestive Heart Failure (UNLOAD) trial, early UF was compared with standard I.V. diuretic therapy in patients with AHF.^[393] UF produced greater weight loss and net fluid loss over 48 hours but no difference in dyspnea scores, creatinine level, or length of stay. There was an associated decrease in HF rehospitalization at 90 days but other important clinical outcomes were not affected. The studies on UF were not powered to address major clinical outcomes, and no long-term evaluation (> 90 days) of the effect on HF or all-cause hospitalizations has been performed.

In a randomized trial Caridorenal Rescue Study in Acute Decompensated Heart Failure (CARRESS-HF) of UF in decompensated HF with CRS, stepped pharmacologic therapy using I.V. diuretics and other medications in an organized fashion was reported to be superior to UF for the preservation of renal function, with similar weight loss. A higher percentage of patients in the UF group had a serious adverse event.^[394]

The potential risks associated with UF include hypotension, bleeding, hemolysis, catheter-related complications, allergic reactions, air emboli, and worsening renal function. Currently, patients receiving UF require systemic anticoagulation, placing an additional risk of systemic bleeding. Estimates regarding the safety of UF, on the basis of the published literature, are from small RCTs and cannot be readily extrapolated to a broader population and in centres without experience and expertise in UF.

7.3.4 Sleep apnea

7.3.4.1 Sleep disordered breathing in HF

The subject of sleep apnea, generally referred as sleep disordered breathing (SDB) in patients with HF has recently been extensively reviewed.^[395] There are 2 types of SDB, namely obstructive sleep apnea (OSA) and central sleep apnea (CSA), which operate through different pathophysiological mechanisms, although they can interact with each other.^[396] OSA results from collapse of the pharynx. Struggling to breathe causes generation of negative intrathoracic pressure, leading to loading of the ventricles. CSA results from either a reduction in central respiratory drive or instability in feedback control of the central respiratory centre. It might be a consequence of HF, but when present, increases the risk of arrhythmias and worsen prognosis.^[396]

About 40% of patients with HF have CSA and 11% have OSA.^{[396]–[398]} A significant number patients with HF with SDB remain undiagnosed, possibly because of a lack of awareness and limited availability of sleep laboratories. Although nocturnal polysomnography in a sleep laboratory is the preferred diagnostic method,^[399] home-based unattended sleep studies have been used as screening tools.^[396] Patients with HF should be asked screening questions, such as on snoring and falling asleep during the day.

7.3.4.2 Treatment of SDB

There have been several RCTs of CPAP therapy in patients with HF and OSA or CSA. For OSA, small studies consistently showed that CPAP reversed obstructive apnea and improved oxygenation at night.^[396] There are also small RCTs of CPAP in CSA, when applied for at least 1-3 months, reduced the apnea-hypopnea index and improved EF and functional class.^[396] One study reported a reduction in combined mortality and heart transplantation rate in HF with CSA, but not in patients without CSA,^[400] which in part prompted the conduct of the Canadian Continuous Positive Airway Pressure for Patients With Central Sleep Apnea and Heart Failure (CANPAP) trial.^[401] This study was terminated early because of a lower than expected event rate. Despite a reported 50% reduction in the apneahypopnea index and 6-minute walking distance, the transplantation-free survival and rates of HF-related hospitalizations were identical in both groups. Thus, data to date suggest that although CPAP alleviated CSA and improved cardiac function, its effectiveness in improving clinical outcomes remains unclear.

The adaptive servo-ventilator has been designed for the treatment of CSA and provides a baseline degree of ventilatory support, in which the patient’s ventilation is servocontrolled to maintain the ventilation at 90% of the long-term average.^[402] Short-term RCTs have shown abolition of CSA but an inconsistent effect on EF.^[395] In the recently published Adaptive Servo Ventilation in Patients With Heart Failure (SERVE-HF) trial,^[403] 1325 patients with an LVEF ≤ 45%, apnea-hypopnea index ≥ 15 events per hour, and a predominance of central events were randomized to receive medical treatment with adaptive servoventilation or medical treatment alone. The primary end point of death, cardiovascular interventions, and HF hospitalization, was not altered. However, all-cause mortality and cardiovascular mortality were significantly higher in the adaptive servo-ventilation group than in the control group. The ongoing Effect of Adaptive Servo Ventilation on Survival and Hospital Admissions in Heart Failure (ADVENT-HF) trial (NCT01128816) might provide more insight into the role of servo-ventilation in HF.

7.4 Acute heart failure

7.4.1 Diagnosis, evaluations, and investigation

The diagnosis of AHF is on the basis of a constellation of symptoms (eg, orthopnea and shortness of breath on exertion) and signs (eg, edema and respiratory crackles) and supported by targeted investigations.^[404],[405] Physical examination evaluates systemic perfusion and presence of congestion.^{[182],[405]–[407]} Laboratory testing, ECG, chest x-ray, and ECHO are all important to obtain relatively efficiently.^[406]

Clinicians should consider potential etiology and precipitating factors (Table 3). In many cases (75%-80%), a precipitant can be found (Table 25).^[408],[409] Failure to uncover the responsible precipitating factor might lead to intractable HF. Noncompliance with diet or medication intake, infections, arrhythmias, pulmonary embolism, and ACS are frequent situations that might cause AHF.

It is essential to perform an ECG in AHF, although it might sometimes be normal. It assists in identifying rhythm abnormalities (AF, flutter, or bradycardia and ventricular tachycardia), ACS,^[410] RV, LV, or atrial hypertrophy or strain, and myopericarditis. Cardiac arrhythmias should be evaluated using a 12-lead ECG and continuous ECG monitoring. A chest x-ray should also be performed in all patients with suspected AHF within the first 1-2 hours of arrival to assess cardiac size and shape, pulmonary congestion, and other pulmonary conditions (Fig. 8).

The utility of NPs to exclude (“rule out”) or confirm (“rule in”) the diagnosis in the appropriate clinical scenario is well established and discussed in section 6. Biomarkers/ NPs. 57,^{[406],[411],[412]} Several clinical scoring systems have been derived and validated and combine commonly used clinical features with NP values to improve diagnosis and decisionmaking (Table 26).^[413],[414] One such clinical scoring system) was developed from the ProBNP Investigation of Dyspnea in the Emergency Department (PRIDE) study.^[413]

7.4.2 Initial and ongoing treatment

Oxygen should be used cautiously in normoxic patients because of concerns of increasing systemic vascular resistance and reducing cardiac output.^[415],[416] There is a paucity of evidence to support the use of I.V. morphine to treat dyspnea, and data suggest there might be adverse effects, even after accounting for the severity of illness, comorbid conditions, and cointerventions (Fig. 9).^{[417]–[419]}

133. We recommend that morphine not be used routinely in patients with AHF (Strong Recommendation; Moderate-Quality Evidence).

BiPAP or CPAP should be considered for patients with a high respiratory rate (eg, > 25 breaths per minute) and persistent systemic arterial hypoxemia despite high-flow oxygen administration.^[420],[421] However, routine use of noninvasive ventilation (NIV) is not advisable. In the Three Interventions in Cardiogenic Pulmonary Oedema (3CPO) trial,^[417] patients with acute pulmonary edema were randomized to standard oxygen therapy, CPAP, or NIV, and followed to the primary end point of 7-day mortality. There was no difference between the 3 arms on 7-day mortality rate and 30-day mortality rates, intubation rates, or admission to an intensive care unit. Therefore NIV should be used only in patients with acute respiratory distress unresponsive to medical therapy. NIV carries the risk of worsening RHF, hypercapnia, aspiration, and pneumothorax. Endotracheal intubation might be used if less invasive modes of oxygen delivery fail or if the patient is in cardiogenic shock.

Oral and I.V. diuretics remain the mainstay of early therapy directed toward AHF (Table 27).^[422] Diuretics generally lead to excretion of sodium and water, leading to a decrease in extracellular fluid volume, total body water, and sodium. A reduction in cardiac filling pressures, peripheral congestion, and pulmonary edema usually follow.^[423] I.V. loop diuretics might cause an early decrease in right atrial pressure and PCWP. When using high I.V. doses reflex vasoconstriction might occur. In AHF, by normalizing loading conditions, these high doses might reduce neurohormonal activation in the short-term.^[424]

Diuretic therapy may be initiated in the ambulance,^[425] HF clinic, or in-hospital. Combining loop diuretics with thiazides^[426],[427] or spironolactone^[428] has been proposed and appears effective, with fewer side effects than a higher dose of a loop diuretic. In patients with severe edema, oral loop diuretics might not be adequately absorbed and might be of little use.^[406] Using stepped pharmacologic diuretic therapy is a useful approach and has been used as the control arm in the CARRESS-HF trial (Fig. 10).^[394]

The Diuretic Optimization Strategies Evaluation (DOSE) trial enrolled 308 patients with AHF and tested 2 I.V. strategies (high- vs low-dose furosemide; continuous infusion vs bolus intermittent dose) for the primary end point of global symptom assessment and creatinine at 72 hours.^[422] There was no difference between the continuous infusion and bolus dosing in either symptoms or renal function. There was greater early symptom improvement with high- compared with low-dose diuretics without a difference in renal function. A number of secondary end points favoured high-dose: a greater diuresis, more weight loss, and a lower NT-proBNP resultant level. A systematic review of 9 additional small trials showed similar findings (Fig. 11).^[429] Thus, there is no advantage in the routine use of continuous diuretic infusions and a higher dose of diuretics could be considered for many patients, with careful observation of renal function and electrolytes. Diuretic responsiveness is another consideration and metrics for the re-evaluation of a patient’s individual response have been reported.^[425],[431] For example, weight loss (or urine output) per diuretic dose unit have been retrospectively evaluated to identify patients with a poor response to diuretics (over 1 day losing < 0.4 kg per 40 mg furosemide) as those at higher risk for short- and long-term morbidity.^[430]

Vasodilators have not been shown to convincingly reduce mortality or reduce rehospitalization rates.^[432] I.V. isosorbide dinitrate (in conjunction with low-dose furosemide) was tested against low-dose nitrates with high-dose diuretics.^[433] This prehospital trial of 110 patients showed that the strategy of high-dose nitroglycerin (compared with high-dose I.V. diuretics) reduced mechanical ventilation rates, and improved oxygen saturation. The Vasodilatation in the Management of Acute CHF (VMAC) trial compared nesiritide, nitroglycerin, or placebo combined with standard therapy for 3 hours, followed by nesiritide or nitroglycerin combined with standard treatment for 24 hours in AHF.^[434] The primary end points of changes in PCWP and patient self-evaluation of dyspnea at 3 hours were improved with nesiritide vs placebo. However, nitroglycerin improved early, short-term dyspnea assessment compared with placebo. The Acute Study of Clinical Effectiveness of Nesiritide in Decompensated Heart Failure (ASCEND-HF) trial tested nesiritide vs placebo in 7141 patients with AHF enrolled within 24 hours of first I.V. medication.^[435] Nesiritide did not reduce mortality, rehospitalization, or the composite of these end points at 30 days. Dyspnea was modestly improved at 6 and 24 hours. The use of nitroprusside in AHF has not been supported by adequately powered RCTs. However, observational studies support its use in advanced HF by clinicians with experience and expertise in managing low-output acute or sub-AHF.^[436] I.V. serelaxin, a vasodilator, was tested in the Preliminary Study of Relaxin for the Treatment of Acute Heart Failure? (PRE-RELAX) trial of 234 patients and had a modest improvement in dyspnea compared with placebo. The Relaxin for the Treatment of Acute Heart Failure (RELAX-AHF) trial tested serelaxin vs placebo in 1161 patients with AHF with a SBP > 125 mm Hg and enrolled within 16 hours of attending the ED. Serelaxin reduced dyspnea measured using a visual analogue scale over 5 days but not using a Likert scale over 24 hours. There was no reduction in the secondary end points of cardiovascular death, hospitalization for heart or renal failure, or days alive out of hospital up to 60 days, but a reduction in mortality at 180 days was seen. There are ongoing trials of serelaxin. There was no clinically meaningful improvement in outcomes with early upfront use of ularitide or TRV-027.^[437],[438]

Inotropic agents have not been shown to improve patient outcomes.^{[406],[439]–[441]} The Outcomes of a Prospective Trial of Intravenous Milrinone for Exacerbations of Chronic Heart Failure (OPTIME-CHF) trial randomized 951 patients admitted for HF to a 48-hour infusion of milrinone or placebo.^[440] New-onset atrial arrhythmias, worsening HF, and symptomatic hypotension requiring intervention occurred more frequently in the milrinone group. A nonsignificant increase in the number of deaths in-hospital and after 60 days was seen in the milrinone group. A post hoc analysis showed a higher incidence of death or rehospitalization in patients with underlying ischemic HF etiology.^[441] Trials using levosimendan have not shown additional benefit compared with placebo^[442]; omecamtiv mecarbil is undergoing further testing in a RCT.^[443] Low-dose dopamine has been studied in the context of AHF^[444],[445] and does not improve clinical symptoms, renal function, or reduce clinical events.

Continuation of a β-blocker upon admission for AHF is considered safe on the basis of the limited data available, including patients receiving inotropes.^[446],[447] In an RCT of 169 patients with AHF, patients either discontinued β-blockade for 3 days or continued the medication unchanged. The trial showed that continuing the β-blocker was noninferior for the primary end point of dyspnea and wellbeing and was associated with a higher rate of β-blocker prescription at 3 months.^[447]

Vasopressin receptor antagonists (eg, tolvaptan) can rapidly and effectively reduce body weight and restore serum sodium in patients with significant symptomatic hyponatremia with hypervolemia and congestion.^[448],[449] The use of a vasopressin antagonist has not yet been associated with mortality or rehospitalization reduction.^[450] A subgroup of 11% and 3% of the patients in the Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study With Tolvaptan (EVEREST) trial had a serum sodium < 135 mmol/L and < 130 mmol/L respectively, and in post hoc analysis, the latter patients had an association with fewer clinical events when treated with tolvaptan.^[451]

7.4.3 Initial and ongoing monitoring and disposition decisions

The extent of monitoring will depend on the disease severity and the response to therapy.^[406] Vital signs (including BP, heart rate, O2 saturation, and daily weight) should be measured on a regular basis until stabilization. Laboratory tests should be repeated regularly (eg, daily in the first 2-3 days): electrolytes, creatinine, and complete blood count, if abnormal. Electrolyte abnormalities, especially hypokalemia, which has been linked to ventricular arrhythmias,^[452] should be prevented or corrected promptly. There is limited evidence to support measurement or replacement of magnesium in patients with AHF. Significant renal impairment might require more frequent laboratory testing.

Decisions to admit a patient from the ED to hospital are complex and require integration of the patient’s clinical stability and preferences and health system features including the availability of appropriate outpatient follow-up. In Canada, 60%-80% of patients in the ED are admitted with AHF.^[446] Whereas there are HF risk models to determine overall mortality risk, there are few instruments to guide clinicians which patients need admission to hospital, continued observation, or discharge home with close follow-up.^[453] One instrument developed from prospectively collected data on 1033 patients with AHF created a nomogram for predicting 5- and 30-day clinical events.^[454] Using different risk of future event thresholds, the nomogram identified a group of patients potentially eligible for safe ED discharge. Two other Canadian studies have developed prognostic risk scores that are as yet tested in RCTs. Table 28 highlights key considerations for these decisions; also see Table 29.

Clinical deterioration despite initial therapy requires closer supervision, such as transfer to an intensive care unit.^[455] Patients in cardiogenic shock or those who have difficulty voiding should have a urinary catheter to monitor urinary output; however, recording of “ins/outs” (also known as fluid balance) is not necessary in all clinically stable patients and can require significant resources to obtain accurately.^[406] The decision to insert an arterial line depends on the need for either continuous analysis of BP because of hemodynamic instability or the requirement for repeated arterial blood gas analyses.^[406] The use of a central I.V. line depends on the need for delivery of fluids and drugs or for monitoring central venous pressure and oxygen saturation. However, in the critically ill, right atrial pressure does not correlate well with left-sided filling pressures.^[456] The insertion of a pulmonary artery catheter is not usually necessary for making a diagnosis or ongoing management of AHF.^{[457]–[459]} It might, however, be useful to distinguish between cardiogenic and noncardiogenic shock, to guide therapy in the presence of severe diffuse pulmonary disease, or in hemodynamically unstable patients who do not respond in a predictable fashion to therapy.^[460]

7.5 Special circumstances

7.5.1 Cardiomyopathies

7.5.1.1 HCM

HCM is a disease of the myocardium characterized by pathological and disproportionate hypertrophy of the left ventricle and, sometimes, the right ventricle. It is most often an inherited condition with an autosomal dominant pattern with variable penetrance. The onset of HCM can occur in early as well as late adulthood,^[461] with a risk of sudden death, often at a young age. Most patients with HCM are asymptomatic.^[462] Symptoms and signs might include chest pain, dyspnea, palpitations, or syncope. Symptoms can arise as a consequence of LV outflow tract obstruction (with secondary MR), tachyarrhythmia (especially ventricular tachycardia and AF), and myocardial ischemia; chest pain is common, even in the absence of CAD^[463] and impaired diastolic function or systolic dysfunction. In a small percentage of cases, the disease progresses to a burnt-out phase, with the development of severe systolic dysfunction that resembles dilated cardiomyopathy (Table 30).^[463],[464]

HCM should be suspected in any individual who presents with unexplained ventricular hypertrophy, heart murmurs (from dynamic outflow obstruction), abnormal ECG patterns (pseudoinfarction, giant negative T waves), unexplained syncope (particularly in young athletes), or a positive family history. A number of hereditary syndromes (including Friedreich ataxia,^[465] and LEOPARD syndrome^[466]) have also been linked to HCM. Approximately 30% of the cases are diagnosed de novo in elderly patients.^[467]

Transthoracic echocardiography (with or without contrast) is the imaging modality of choice, and might include a provocative test for dynamic LV outflow tract obstruction. CMR imaging may be used if initial imaging results are nondiagnostic and might provide other structural details.^[468] Tissue Doppler imaging might aid in early diagnosis.^[469] Localized hypertrophy of the interventricular septum is the most common pattern of hypertrophy, but other patterns are also possible. Dynamic LV outflow obstruction occurs in approximately 30%-50% of cases and might be latent.^[470] Differentiation must be made from other pathologies that mimic HCM in appearance, including concentric hypertrophy due to systemic hypertension, physiological hypertrophy seen in trained athletes, and discrete/disproportionate upper septal hypertrophy in elderly patients.

7.5.1.2 Restrictive cardiomyopathy and constriction

Restrictive cardiomyopathy (RCM) as the etiology of HF can be difficult to recognize and is the least common category among the cardiomyopathies.^[471],[472] It is characterized by myocardium with markedly stiff ventricular walls, restrictive ventricular filling, and reduced diastolic volume of either or both ventricles, and normal or near-normal systolic function.^[472],[473] Myocardial fibrosis, infiltration, or endomyocardial scarring is responsible for the diastolic dysfunction.^[473] Consequently, RCM shares similar functional characteristics with constrictive pericarditis, and differentiation between the two might be difficult but imperative because surgery might potentially cure the latter.^[471] Amyloidosis is a frequent cause, but many other situations might also result in RCM. Rare hereditary forms of RCM have been described, such as troponin I gene mutations or in association with skeletal muscle disease (Table 31).^[473],[474] Prognosis of patients varies substantially, but it is generally one of inevitable downward symptomatic progression with a high mortality.^[475]

7.5.1.2.1 Specific imaging and diagnostic tests for restrictive cardiomyopathies and constriction

There are a number of selected tests that will aid either the diagnosis, prognosis, or a selection of appropriate therapy for patients with a suspected RCM or constriction. These are presented and discussed in the following sections. Not all laboratory or imaging tests are readily available in all regions, and therefore, expertise should be sought where necessary.

7.5.1.2.2 Physical examination and ECG

Patients usually exhibit typical symptoms and signs of HF, including Kussmaul sign that might be disproportionate to the degree of systolic or valvular dysfunction. Hepatomegaly, ascites and, in more advanced cases, anasarca, might be present. Notably, the apex beat is usually palpable in RCM, but not in constrictive pericarditis.^[473] The ECG is frequently abnormal with decreased voltage intraventricular conduction delay, or poor R wave progression mimicking MI.^[476] Sinus node disease is frequent, and typical characteristics of the sick sinus syndrome are encountered.^[477] Arrhythmias, mostly AF, are frequent and might be caused by amyloid deposition.^[478]

7.5.1.2.3 Laboratory findings

Specific morphological features (especially musculoskeletal) in the setting of renal disease and a family history of metabolic abnormalities suggest the presence of an infiltrative disorder, such as Fabry disease. Fabry disease is a rare X-linked disorder that can present with neuropathy, renal failure, and HF in affected men and hemizygous women. Genetic testing can determine the exact gene defect, but elevated urinary globotriaosylceramide, and reduced plasma a-galactosidase activity are diagnostic.^[479] These disorders usually require tissue diagnosis, either from cardiac tissue or other affected organs, such as the kidneys.

Abnormalities in serum protein electrophoresis suggest amyloidosis, whereas a high plasma level of ferritin, in combination with increased transferrin saturation, suggests hemochromatosis. Ventricular arrhythmias might be a forerunner for sudden death, and a signal-averaged ECG has been suggested to help in identifying patients at risk.^[480]

An abdominal fat aspirate is safe and might assist in the diagnosis of amyloidosis. If the result of an abdominal fat aspirate is negative, EMB might help if cardiac amyloidosis is suspected, and kidney biopsy might also be useful if the eGFR is low, suggesting renal involvement. Immunohistochemical staining should be performed because it helps to differentiate between systemic senile, familial, and primary forms of amyloidosis, and clarifies prognosis and management, which differ in the various forms.^[481],[482] Unsuspected hereditary amyloidosis might be present in nearly 10% of patients thought to have the primary amyloid light-chain form.^[483] The absence of concomitant symptoms and signs of systemic illness, nonspecific laboratory findings, and negative EMB strongly suggest idiopathic RCM.

7.5.1.2.4 Imaging

Echocardiography might be normal early on, but small ventricular chambers with increased wall thickness, large atria, and thickening of valvular apparatus and interatrial septum are often seen. LV systolic function is usually normal, but might be reduced in advanced disease. Pericardial effusion is common, but rarely results in tamponade.^[484] The sparkling appearance of the thickened walls, presumably related to amyloid deposition, was previously believed to be typical on echocardiography, but is less reliable compared with recent technology.^[485] Doppler echocardiography is useful for the diagnosis, prognosis, and should be evaluated carefully in centres with experience and expertise.^[486],[487]

CMR can precisely image functional, morphological changes,^[488] fibrosis,^[489] and inflammation^[490] and might provide early insight into the disease process and etiology. CMR is useful in amyloidosis,^[491] with subendocardial contrast enhancement associated with nonsuppressible signal of remote myocardium. CMR might increase the sensitivity of myocardial biopsy,^[492] and might help to direct treatment.^[493] CMR might be helpful in distinguishing between active myocardial disease and clinical remission in systemic lupus erythematosus,^[494] documenting the presence of cardiac fibrosis in Churg-Strauss syndrome, a rare form of systemic vasculitis,^[495] and detecting early indications of iron overload in thalassemia patients.^[496]

Scintigraphy with technetium-99m pyrophosphate, as well as with other agents that bind to calcium, is frequently positive, with extensive amyloid infiltration. Scanning with specialized agents might also identify sympathetic denervation in patients with cardiac amyloidosis.^[497]

7.5.1.2.5 Hemodynamic findings

The characteristic hemodynamic feature of RCM as well as constrictive pericarditis on cardiac catheterization is a rapid early decrease in ventricular pressure at the beginning of diastole, with a rapid increase to a plateau in early diastole (this latter finding might be absent in RCM)dthe ‘dip and plateau’ or ‘square root’ sign. Systemic as well as pulmonary venous pressures are elevated (frequently > 50 mm Hg in RCM, although lower in constrictive pericarditis), and patients with RCM typically have LV filling pressure that exceeds RV filling pressure by more than 5 mm Hg (this difference can be revealed by exercise, fluid challenge, or the Valsalva manoeuvre). In constrictive pericarditis, filling pressures are similar in both ventricles, and the plateau of the RV diastolic pressure is usually at least one-third of the peak RV systolic pressure, but is frequently lower in RCM. In the setting of rapid early diastolic ventricular filling, an increase in RV peak systolic pressure during inspiration might occur with a reduction of LV peak systolic pressure.

Differentiating between RCM and constrictive pericarditis might be difficult and thus multiple modalities in centres experienced with these techniques is important. EMB and CT might also be useful for differential diagnosis; however, in rare cases, open biopsy might be required.^[471]

7.5.2 Ethnicity

On the basis of a Statistics Canada 2006 census,^[498] the visible minority population surpassed 5 million, reaching 16.2% of the population. In Ontario, more than 1.5 million people are of Chinese, South Asian, black, or Aboriginal descent. To understand and manage a person’s illness it is necessary to appreciate the effects of the person’s culture and social environment. This is perhaps most relevant in the health care management of minority groups. Morale is crucial to the patients’ adaptation and their maintenance of involvement in their management; miscommunications as a result of ethnocultural differences might have a detrimental effect on their adaptation to their illness. Furthermore, health care providers might contribute to the ethnic care disparities through clinical uncertainty and stereotyping of health behaviours related to minority patients.^[499] However, the use ethnicity as a way to differentiate patients might be debatable and differential medical treatment on the basis of the colour of one’s skin has been associated with detrimental outcomes for ethnic minorities.^[500] A significant contributor is the paucity of research to clearly identify the sources of these differences in outcomes in ethnic groups and to distinguish among biological, environmental, or social causes of disease differences.^[501] Evaluation of disease differences in subsegments of the population is needed to understand the mechanisms of pathophysiology and to optimally target therapeutic responses. Thus, effective research that would contribute to a reduction in health care disparities requires collection of data on health status in ethnic populations and assessment of differences in disease patterns. It also requires clinical trials to include adequate numbers of diverse populations to probe for differences in pathophysiology including environmental or social factors contributing to disease and responses to treatment. Where differences are observed among population segments, clinical trials focused in these population groups are warranted.^[501],[502]

7.5.2.1 General considerations

There have been very few published population-based epidemiological studies or large-scale randomized controlled studies of HF in countries outside North America, Europe, and Australia,^[503] regions from which most of the minority groups that reside in the Western countries emigrated. For example, it is generally believed that rheumatic heart disease and congenital heart disease remain important causes of HF in sub-Saharan Africa and certain parts of Asia and South America. Hypertension is thought to be an important cause of HF in Asia, and in the African and African American population, whereas Chagas disease is an important etiology in subjects from South America.^[504] Although it is useful to remember region-specific etiologies of HF particularly when managing recent immigrants from the regions where the minority groups were resident, it should be remembered that because these regions also constantly undergo epidemiological and economic transitions the epidemiology of HF is likely to be increasing similar to that of the Western world. The INTERHEART study has shown that the effect of conventional and potentially preventable risk factors on the risk of MIs are consistent across different geographical regions and different ethnic groups.^[505] This implies that simple measures that can prevent MI and likely the subsequent development of HF are equally applicable to different ethnic populations in different geographic locations. The recently published Cardiovascular Health in Ambulatory Care Research Team (CANHEART) Immigrant study^[506] shows that most immigrant groups in Canada have lower rates of major cardiovascular events than long-term residents of similar age; striking variations in the event rates exist between immigrants from different ethnic background. East Asian immigrants, predominantly of Chinese descent, had the lowest burden of risk factors and events overall, although the event rate increased with greater duration in Canada. There are also high-risk groups such as South Asian immigrants, who had a high burden of traditional risk factors and frequent cardiovascular events. In general, patients in the Asia Pacific region have historically less CAD as etiology, onset at younger age, fewer uses of devices, more diabetes, and more uses of parenteral agents during acute episodes.^[507]

There is little evidence to indicate that criteria used to diagnose HF differ between ethnic populations. For example, a recent study from the United States has shown that the diagnostic performance of the biomarker N-terminal BNP is similar in African American and non-African American individuals.^[508] Evidence that different ethnic groups have the same mortality benefit from current standard therapy is scant because very few large RCTs have included regions outside Europe and the United States. There have been smaller trials that confirmed the effectiveness of ACEis and β-blockers in patients from Africa and Asia.^[504],[509] Because of the fundamental nature of the derangements in HF, it is likely that the current treatment approach such as blockade of neurohormonal activation and the judicious use of devices will be similarly effective, although one cannot rule out the possibility that the degree of response to treatment might vary among ethnic groups. Details that pertain to specific minority populations are available in section 7.5.2.2 of the Supplementary Material. Recommendations and practical tips on the management of patients with HF from the 4 largest ethnic minority groups in Canada are shown in Table 32.

7.5.3 Pregnancy

HF during pregnancy has a number of potential etiologies including PPCM. Preconception counselling is recommended in women with inheritable and/or a known history of HF or PPCM. Maternal risk assessment and frequency of expert follow-up should be scheduled as recommended by the modified WHO risk classification.^[510],[511] High-risk patients should be referred to a multidisciplinary team with expertise in the management of HF and highrisk pregnancy.^{[512]–[514]}

7.5.3.1 Diagnosis and management

Hemodynamic changes in normal pregnancy can precipitate HF in patients who are susceptible or have another underlying etiology (Table 33). Decompensation can occur at any time; however it is more common in the late stages of pregnancy and peripartum. Physical examination for HFrelated physical signs can be challenging in pregnancy. It is important for clinicians to recognize cardiovascular symptoms and signs that are not normally present in pregnancy (Table 34). Echocardiography remains the preferred imaging modality for HF during pregnancy. Women with AHF during pregnancy should be managed according to the guidelines for AHF and should be referred to a tertiary care centre with expertise in advanced HF management, including MCS and cardiac transplantation.^[515],[516]

7.5.3.2 Medical therapy of HF in pregnancy

HF in pregnancy should be treated according to the CCS HF guidelines for acute and chronic HF. However, many standard therapies of HF are considered fetotoxic and should not be used during pregnancy including ACEis, ARBs, and MRAs.^[517],[518] There are no data on the use of ARNIs and thus they are not recommended during pregnancy. When β blockade is used, β1-selective drugs (eg, metoprolol) should be preferred. Atenolol should not be used.^[519] Diuretics can be used if pulmonary congestion is present. Medications that might be used for pregnant women with HF are shown in Table 35. An additional comprehensive list of medications is available at www. motherrisk.org.^[515],[516]

7.5.3.3 PPCM

PPCM is a diagnosis of exclusion typically defined as HF with an LVEF of < 45% within 1 month before delivery to 5 months postpartum. Pathophysiologic mechanisms are not clearly defined but might include genetic predisposition,^[520] oxidative stress, and immune mechanisms,^[521],[522] as well as viral infections^[521] and prolactin.^[523] Risk factors for the developments of PPCM include multiparity and multiple fetal gestation, advanced maternal age, family history, ethnicity, hypertension, preeclampsia, smoking, diabetes, and prolonged tocolytic therapy.^[524]

In addition to standard diagnostic tests for HF and pregnancy, there is now data to support the use of biomarkers for diagnosis as well as prognosis in PPCM.^[525],[526] There have been several case reports and a small RCT to evaluate bromocriptine^[527] with inconclusive results and uncertain safety.^{[528]–[530]}

Despite advances in HF treatment mortality as well as morbidity related to PPCM remains high. In case series, LV systolic function returns to normal in 23%-72% of patients.^{[524],[531]–[533]} Less is known about the risk of subsequent pregnancy; however, the 2 largest studies of PPCM suggest relapse occurs in almost one-third of the cases.^[534],[535] Thus most would agree that individualized counselling should occur in individuals with PPCM and they should be advised against future pregnancy particularly if recovery of LVEF has not occurred.^[524],[536]

145. We recommend that echocardiography be performed in women with worsening or suspected new-onset HF during pregnancy (Strong Recommendation; Low-Quality Evidence).

146. We recommend prepregnancy counselling in all women with a known history of HF or PPCM (Strong Recommendation; Low-Quality Evidence).

147. We recommend preconception genetic counselling in women with inheritable cardiac diseases that can affect cardiac function, including inheritable cardiomyopathies (Strong Recommendation; Low-Quality Evidence).

148. We recommend maternal risk assessment and frequency of expert follow-up should be determined using the modified WHO risk classification (Strong Recommendation; Low-Quality Evidence).

149. We recommend that decisions regarding timing and mode of delivery should be on the basis of obstetrical factors (Strong Recommendation; Low-Quality Evidence).

7.5.4 Cardio-oncology and HF

Cancer therapy might result in subclinical and clinical LV dysfunction. The CCS recently published guidelines for evaluation and management of cardiovascular complications of cancer therapy.^[55] Specific recommendations, from that document, related to the development and treatment of LV dysfunction are presented in this document for ease of review; the content has not been modified. Readers are directed to the CCS guidelines for evaluation and management of cardiovascular complications of cancer therapy for a discussion of the evidence underpinning these recommendations. Our current understanding of agents associated with LV dysfunction, their mechanism, and time course for toxicity as well as patient- and treatment-related risk factors are summarized in Table 36. A comprehensive table of cancer treatments associated with all forms of cardiotoxicity are published elsewhere.^[55] Additional information is also presented in section 7.5.4 of the Supplementary Material.

7.5.5 Myocarditis

Myocarditis, which often presents as HF, is an inflammatory process affecting the myocardium as a result of external antigen triggers such as viruses, bacteria, parasites, drugs, and others, or internal triggers such as autoimmune reaction to self-antigens. The WHO definition of myocarditis is on the basis of established histological, immunological, and immunohistochemical criteria.^[537],[538]

The incidence of myocarditis is not well established because many patients with clinically suspected myocarditis do not undergo EMB or CMR. In patients with unexplained nonischemic cardiomyopathy, biopsy proven myocarditis was shown in 9%-16% of cases.^[539],[540] The prognosis for patients with myocarditis is usually favourable. One prospective study suggested that approximately one-third of patients who present with acute myocarditis do not develop HF, one-third develop ventricular dysfunction with subsequent recovery, and approximately one-third are left with significant ventricular dysfunctionda small subgroup of subjects progressively deteriorate and require significant support (MCS or cardiac transplantation).^[541],[542]

Clinical presentations of patients with myocarditis ranges from asymptomatic patients with abnormal ECG or ECHO findings to patients with atypical symptoms of fatigue, palpitations, chest pain at rest, and patients with arrhythmias, HF, cardiogenic shock, or sudden death. A high index of clinical suspicion along with biomarkers and noninvasive investigations like ECHO and CMR are necessary to make a diagnosis.

Markers of myocardial necrosis (troponin I or T) can assist in the diagnosis of myocarditis but a low or normal troponin level does not rule out myocarditis. The sensitivity of an elevated troponin I level in biopsy proven myocarditis in the Multicentre Myocarditis Treatment trial was 34% and specificity was 82%.^[543] The ECG findings might include arrhythmias (ventricular or supraventricular), AV block, pattern of acute injury or pericarditis, nonspecific repolarization abnormalities or, rarely, might be normal. Echocardiographic findings might include segmental or global LV dysfunction, RV dysfunction, or pericardial effusion.^[544]

CMR is the most important noninvasive investigation in the diagnostic workup of myocarditis. CMR allows for accurate and quantitative assessment of LV morphology, volumes, as well as global and regional ventricular function. More importantly however, CMR can be used to detect inflammation, by allowing visualization of myocardial hyperemia (early gadolinium enhancement), intracellular and interstitial edema (water-sensitive CMR), and necrosis/fibrosis (LGE) imaging. The combined protocol (Lake Louise criteria) has high specificity of up to 91% and a sensitivity of 67%. LGE imaging alone, however, also has a variable sensitivity of 44%- 100%^[545] and is also not specific for acute vs chronic/healed myocarditis.^[546]

In a study of 104 patients with subacute myocarditis, the extracellular volume with LGE imaging showed improvement in the diagnostic accuracy of CMR compared with standard Lake Louise criteria.^[545] Myocardial mapping as a novel CMR approach might further improve diagnostic accuracy^[547],[548]; further clinical research is ongoing.

The yield of EMB to provide clinically relevant information using Dallas criteria (histological criteria) is relatively low. Using immunohistochemical criteria and viral polymerase chain reaction in addition to histological criteria increases the diagnostic value of EMB.^[549],[550] EMB should be considered if the results of biopsy are likely to result in a change of patient management. Giant-cell myocarditis is an example in which biopsy results provide information on prognosis and the need for immunosuppressive treatment. Other situations in which EMB provides unique information are disorders associated with sarcoidosis, hypereosinophilia, infiltrative cardiomyopathies, and others.

EMB is indicated for patients who present with an unexplained new onset of HF (< 2 weeks) and hemodynamic compromise. In this scenario, EMB might help establish diagnosis but has known safety considerations and diagnostic yield issues. EMB is also indicated in patients with new-onset HF associated with ventricular arrhythmias, high-degree AV blocks, and in patients who failed to respond to medical therapy. Ventricular arrhythmias and high-degree AV blocks are frequently associated with giant-cell myocarditis or sarcoidosis.^[551]

Irrespective of clinical presentation, patients with myocarditis and HF should be treated with standard medical therapy for HF. The routine use of immunosuppression in myocarditis is not recommended. There are several small RCTs investigating the use of steroids alone or in combination with cyclosporine or azathioprine vs placebo in patients with myocarditis. In the largest study, 111 patients with biopsy proven myocarditis and reduced LV function (EF < 35%) were randomized to conventional therapy vs prednisone and either azathioprine or cyclosporine. At 12 months there was no survival benefit and no significant difference in LV function improvement.^[539] In a study of 85 patients with biopsy proven virus-negative myocarditis, patients were assigned to placebo or prednisone in combination with azathioprine. At 6 months 88% of patients in the prednisone/azathioprine group showed significant improvement in echocardiographic parameters.^[552] There are no large, randomized, placebocontrolled studies investigating antiviral therapy in patients with myocarditis. A high dose of I.V. immunoglobulin was evaluated in a small number of patients with myocarditis and the treatment was ineffective.^{[553]–[555]}

Patients with known or suspected myocarditis should be referred to a centre where expertise in the diagnostic assessment and treatment of myocarditis is available. The urgency of referral is dependent on the clinical course. Reports of small case series suggest that aggressive medical support, including a ventricular assist device, along with medical therapy, has allowed for eventual ventricular function recovery and device explantation without the need for transplantation.^[556],[557]

The intensity of follow-up for patients with myocarditis is dictated by the extent of cardiac dysfunction, the severity of the clinical presentation and the response to therapy. Followup consists of ongoing clinical assessment and might include echocardiographic assessment of cardiac function or CMR evaluation of ongoing inflammation. Emerging evidence suggests that CMR follow-up might be useful for predicting outcomes.^[558] Persistent inflammation at 4-week follow-up indicates a worse prognosis in patients with no further inflammation. Patients with a positive clinical response to therapy, including improvement in or normalization of cardiac dysfunction, should also undergo clinical follow-up within approximately 3-6 months to confirm clinical stability. Patients with a continued or worsening course, which will be dictated by the clinical severity of symptoms and LV dysfunction, require ongoing expert follow-up.