The final module of the CHF series addresses three areas that require integrating the pharmacological knowledge built across previous modules into complex, real-world clinical scenarios. First, HFpEF, a syndrome that accounts for at least half of all heart failure presentations, has a distinct pathophysiology from heart failure with reduced ejection fraction (HFrEF), and has historically resisted pharmacological intervention, is now entering a new era with the emergence of sodium-glucose cotransporter 2 (SGLT2) inhibitor evidence and evolving understanding of its heterogeneity. Second, the interface between pharmacological therapy and cardiac device therapy (implantable cardioverter-defibrillator, cardiac resynchronization therapy, left ventricular assist device) creates unique drug interaction and management challenges that the clinician must navigate. Third, heart failure in special populations, specifically patients with chronic kidney disease and patients who are pregnant, requires thoughtful modification of the standard pharmacological approach, balancing the benefits of guideline-directed medical therapy (GDMT) against population-specific risks.
The repeated failure of pharmacological interventions proven effective in HFrEF to demonstrate benefit in HFpEF is not coincidental: it reflects fundamental pathophysiological differences between the two syndromes that have direct pharmacological implications.1 HFpEF is driven not by cardiomyocyte loss and adverse remodeling from neurohormonal activation (the dominant HFrEF pathway), but by a systemic pro-inflammatory state generated by the chronic comorbidity burden (obesity, hypertension, diabetes, metabolic syndrome, sleep apnea, CKD) that activates microvascular endothelial inflammation, drives cardiomyocyte hypertrophy and stiffness through titin hypophosphorylation, and impairs myocardial relaxation.1 Additionally, many HFpEF patients have a component of chronotropic incompetence, exercise-induced pulmonary hypertension, and impaired skeletal muscle oxygen utilization: features not well addressed by renin-angiotensin-aldosterone system (RAAS) or sympathetic nervous system (SNS) blockade. The key pharmacological corollary: drugs that target RAAS/SNS overactivation, which are the cornerstones of HFrEF therapy, do not address the principal drivers of HFpEF, explaining their serial failures in this phenotype.
A sequence of trials applying HFrEF-proven drug classes to HFpEF demonstrated either neutral or harmful results: ACE inhibitors and ARBs: The CHARM-Preserved trial (candesartan in HFpEF, left ventricular ejection fraction (LVEF) >40%) showed a non-significant reduction in cardiovascular (CV) death or HF hospitalization (HR 0.89; p=0.051).2 The PEP-CHF trial (perindopril in elderly HFpEF) similarly showed no significant primary outcome benefit. The I-PRESERVE trial (irbesartan in HFpEF) showed absolutely no benefit. Neither ACEi nor ARBs are guideline-recommended for mortality reduction in HFpEF, though they may be used for coexisting conditions (hypertension, CKD, post-MI).3
Beta-blockers: No large randomized trial has demonstrated mortality benefit from beta-blockers in HFpEF. They are used for coexisting indications (rate control in AF, post-MI, hypertension) but carry a theoretical concern about reducing the compensatory heart rate reserve that HFpEF patients use to augment cardiac output during exercise (chronotropic reserve). Beta-blockers should generally not be added to HFpEF solely for HF benefit.3 Spironolactone (TOPCAT): As reviewed in CHF-04, TOPCAT did not meet its primary endpoint overall. Subgroup analyses and post-hoc data suggest possible benefit in biomarker-confirmed HFpEF, but the evidence is insufficient for a Class I recommendation.2 Spironolactone carries a Class IIb recommendation in selected HFpEF patients (those with elevated BNP, recent HF hospitalization, controlled BP, eGFR ≥30, and K⁺ <5.0 mEq/L).3 Sacubitril/valsartan: PARAGON-HF (2019) randomized 4,822 patients with HFpEF (LVEF ≥45%) to sacubitril/valsartan vs. valsartan. The primary composite endpoint (CV death or HF hospitalization) was not significantly reduced (RR 0.87; p=0.059) overall.2 However, women and patients with LVEF in the range of 45–57% showed significant benefit in post-hoc analyses. The FDA approved sacubitril/valsartan for HFpEF in a specific indication covering the lower end of the HFpEF LVEF spectrum: but guidelines do not give this a strong recommendation (Class IIb for LVEF below normal).3
SGLT2 inhibitors: As established in CHF-05, EMPEROR-Preserved and DELIVER are the landmark trials. The class is the only pharmacological intervention with consistent, robust, replicable efficacy across the HFpEF/HFmrEF spectrum. Current 2022 AHA/ACC/HFSA guidelines give SGLT2 inhibitors a Class IIa recommendation in HFpEF (LVEF ≥50%): the strongest pharmacological recommendation available.3 Dapagliflozin 10 mg or empagliflozin 10 mg once daily regardless of diabetes status is the central pharmacological intervention in HFpEF.
Diuretics: Loop diuretics remain the primary tool for symptom management (decongestion). Unlike in HFrEF, diuretics in HFpEF require particular care regarding volume status: HFpEF patients depend on adequate preload to fill a stiff ventricle, and over-diuresis can precipitate a sharp reduction in cardiac output. The target should be euvolemia, not aggressive volume reduction.3 Blood pressure control: Hypertension is both a major causative factor and an active driver of disease progression in HFpEF. Achieving target BP (<130/80 mmHg in most patients with HFpEF and HTN) is one of the most important modifiable therapeutic interventions. Any antihypertensive class can be used; RAAS blockers (ACEi, ARB) are commonly chosen for combined BP control and potential anti-fibrotic effects.3 Heart rate and atrial fibrillation management: AF is present in approximately 40–65% of HFpEF patients at some point in their clinical course, and its impact on HFpEF symptoms is particularly pronounced because the syndrome is heavily dependent on atrial contribution to ventricular filling and on adequate diastolic filling time. Rate control and rhythm control strategies are both appropriate in HFpEF-AF, with emerging evidence from EAST-AFNET 4 that early rhythm control may be particularly beneficial. Beta-blockers, non-dihydropyridine (DHP) CCBs (which are NOT contraindicated in HFpEF as they are in HFrEF), and digoxin are all reasonable rate-control options in HFpEF with AF.3
HFpEF is not a single disease: it is a syndrome umbrella covering multiple pathophysiological phenotypes with distinct therapeutic implications: Obese HFpEF: The most prevalent phenotype in the modern era. Adiposity drives epicardial fat inflammation, pericardial constraint, systemic inflammation, and adipokine-mediated myocardial stiffness. glucagon-like peptide-1 (GLP-1) receptor agonists (liraglutide, semaglutide) have demonstrated significant reductions in HF hospitalization and symptom burden in obese HFpEF patients in the STEP-HFpEF trial, positioning weight loss therapy as a potential disease-modifying intervention in this phenotype.3 Semaglutide (Ozempic/Wegovy) is now a Class IIa recommendation in obese HFpEF (BMI ≥30 kg/m2) in the 2023 updates to guidance. Hypertensive HFpEF: Blood pressure-driven myocardial hypertrophy and fibrosis are the primary drivers. Aggressive BP control is the cornerstone; RAAS blockade is preferred. AF-HFpEF: As above: rhythm and rate management are particularly important, with a lower threshold for rhythm control strategies. CKD-HFpEF: Cardiorenal interactions are prominent; diuretic resistance is common; SGLT2 inhibitors have demonstrated renoprotective effects alongside HF benefit.
Primary prevention ICD is indicated in HFrEF patients with LVEF ≤35% who have NYHA class II–III symptoms and are on maximally optimized GDMT for at least 3 months (with life expectancy >1 year), as these patients are at high risk of sudden cardiac death.3 The pharmacological interaction with ICD therapy operates in two directions: GDMT before ICD: As demonstrated in CHF-03, optimized GDMT (particularly ARNI, beta-blocker, and SGLT2 inhibitors) can substantially improve LVEF: in some cases raising it above the 35% threshold and potentially making ICD implantation unnecessary. Current guidelines explicitly recommend a 3-month GDMT optimization period before ICD implantation in patients with newly diagnosed non-ischemic HFrEF, to identify patients with significant LVEF recovery.3 Ischemic HFrEF patients with LVEF ≤35% and prior MI (≥40 days) may proceed to ICD evaluation earlier. GDMT after ICD: In patients with ICDs, beta-blockers and amiodarone reduce the frequency of appropriate ICD shocks by suppressing ventricular arrhythmias. Beta-blockers are independently effective; amiodarone is used as adjunctive antiarrhythmic therapy in patients with frequent ICD discharges.3
Dofetilide is also used in selected HF patients with ICD for refractory ventricular or atrial arrhythmias; it requires in-hospital initiation due to QTc monitoring requirements and has important drug interactions (inhibited by cimetidine, trimethoprim, verapamil). Amiodarone and digoxin interaction: A clinically significant pharmacological interaction in ICD-HFrEF patients: amiodarone (commonly used for arrhythmia suppression) increases digoxin levels by 50–100% through inhibition of P-glycoprotein. When amiodarone is started in a patient on digoxin, the digoxin dose must be halved empirically and levels rechecked. Failure to do so is a frequent cause of digoxin toxicity.4
CRT (biventricular pacing) is indicated in HFrEF patients with LVEF ≤35%, NYHA class II–IV, sinus rhythm, and LBBB with QRS complex (QRS) duration ≥150 ms (Class I) or non-LBBB with QRS ≥150 ms (Class IIa).3 CRT restores ventricular synchrony, improving LVEF, reducing functional mitral regurgitation, and reducing symptoms. GDMT is not reduced after CRT: CRT does not replace GDMT; the two interventions have independent, additive benefits. Patients who respond to CRT with LVEF normalization ("super-responders") should continue all four GDMT pillars: as established by the TRED-HF paradigm for GDMT discontinuation in recovered HF.8 Post-CRT LVEF recovery and ICD de-implantation: Some CRT responders with LVEF recovery above 35% have been followed without ICD generator replacement at battery depletion. This remains an area of individualized decision-making; no definitive guideline supports routine ICD de-implantation based on LVEF recovery alone. Beta-blocker and optimal AV delay: In patients with cardiac resynchronization therapy with defibrillator (CRT-D) devices, the AV delay programming interacts with the pharmacological AV conduction effects of beta-blockers and digoxin. Device interrogation and AV delay optimization should be reassessed after significant changes in these drugs.
LVADs provide mechanical unloading of the LV, dramatically improving cardiac output and end-organ perfusion in patients with advanced HFrEF (NYHA class IIIB–IV) as a bridge to transplant or as destination therapy.7 The pharmacological environment in LVAD patients is significantly different from conventional HF management: Anticoagulation: All LVAD patients require therapeutic anticoagulation (typically warfarin, with international normalized ratio (INR) target 2.0–3.0) to prevent device thrombosis and thromboembolic stroke, in addition to antiplatelet therapy (aspirin 81–325 mg daily). Maintaining stable INR is critical; drug interactions affecting warfarin (amiodarone, antibiotics altering gut flora, dietary vitamin K variation) require frequent INR monitoring.7
GDMT continuation on LVAD: RAAS blockers (ACEi/ARB or ARNI), beta-blockers, and MRAs should generally be continued or initiated in LVAD patients who can tolerate them. Neurohormonal blockade promotes reverse remodeling of the native myocardium, which may allow LVAD explantation in some patients ("bridge to recovery").7 Blood pressure management: Optimal MAP on LVAD is typically 70–80 mmHg. Hypertension (MAP >90 mmHg) significantly increases stroke risk; antihypertensive therapy (ACEi, ARBs, hydralazine, amlodipine) is often required. Hypotension (MAP <60 mmHg) reduces device output and may indicate suction events. Arrhythmia management: AF and ventricular arrhythmias are common in LVAD patients. Unlike non-LVAD patients, ventricular tachycardia in LVAD patients may be better tolerated hemodynamically (because the LVAD maintains cardiac output independent of ventricular contraction) but remains proarrhythmic and warrants management. Beta-blockers and amiodarone are used for arrhythmia suppression.
CKD and HF frequently coexist and mutually worsen each other: a bidirectional interaction termed the cardiorenal syndrome (CRS).5 Reduced cardiac output lowers renal perfusion pressure, activating RAAS and SNS and promoting renal ischemia (cardiorenal syndrome type 1 and 2). Conversely, CKD itself drives HF through volume overload, anemia, uremic cardiomyopathy, and CKD-associated RAAS activation (renocardiac syndrome types 3 and 4). The pharmacological challenge is that most HF drugs affect renal function (RAAS blockers reduce intraglomerular pressure, loop diuretics can cause volume depletion and pre-renal acute kidney injury (AKI)), and that CKD alters the pharmacokinetics of many HF drugs (reduced renal elimination of digoxin, bisoprolol, and milrinone).
ACEi/ARBs/ARNI: All are beneficial and should not be routinely withheld in HF patients with CKD.3 Accept a creatinine rise of up to 30–35% after initiation or dose increase. A creatinine rise above 50% or potassium >6.0 mEq/L requires dose reduction or temporary hold. Both sacubitril/valsartan and ramipril have been studied in patients with eGFR as low as 25–30 mL/min/1.73m2 and retain benefit.3 In patients with eGFR 15–30 mL/min/1.73m2, start at the lowest available dose and monitor closely. Beta-blockers: Most are hepatically metabolized and do not require dose adjustment in CKD. Bisoprolol (approximately 50% renally excreted) and atenolol (primarily renal) may accumulate in severe CKD; carvedilol and metoprolol succinate are primarily hepatically metabolized and are generally safe without dose adjustment. Beta-blockers are beneficial across the CKD spectrum in HFrEF.3 MRAs: eGFR ≥30 mL/min/1.73m2 is the standard threshold for MRA initiation. Eplerenone is preferred in moderate CKD (eGFR 30–60) based on EMPHASIS-HF data.5 In patients with eGFR 30–45, start at the lowest dose (25 mg every other day) and monitor K⁺ and creatinine weekly for the first month. Patiromer use to manage hyperkalemia can enable MRA use in patients who would otherwise be unable to tolerate them.
SGLT2 inhibitors: As detailed in CHF-05, the HF indication allows use at lower eGFR thresholds than the glucose-lowering indication. Dapagliflozin is recommended for HFrEF and HFpEF without a specified minimum eGFR; empagliflozin with eGFR ≥20. The renoprotective effects of SGLT2 inhibitors (reduction in eGFR decline over time, reduced risk of end-stage renal disease (ESRD)) make them particularly valuable in heart failure with chronic kidney disease (HF-CKD).3 Loop diuretics: Furosemide dose requirements increase substantially as eGFR falls, because drug delivery to the tubular lumen depends on renal secretion. In patients with eGFR <30 mL/min/1.73m2, doses of furosemide up to 200–400 mg daily (oral or IV) may be required to achieve adequate diuresis. Torsemide retains better oral bioavailability and is preferred in advanced CKD. Metolazone retains diuretic activity even at low eGFR and is a useful adjunct in diuretic-resistant HF-CKD.5 Digoxin: Renally excreted (70%); half-life is markedly prolonged in CKD. In patients with eGFR <60, reduce daily dose to 0.0625 mg (62.5 mcg) daily or 0.125 mg every other day; some patients with eGFR <30 can be managed on 0.0625 mg every other day. Check levels frequently and vigilantly for toxicity. Hypokalemia (common in CKD patients on diuretics) dramatically sensitizes the myocardium to digoxin toxicity.4
NSAIDs: Prostaglandins are critical mediators of renal afferent arteriolar dilation and glomerular perfusion, particularly in the HF state where renal circulation is already vasoconstricted. NSAIDs inhibit prostaglandin synthesis, precipitating acute prerenal AKI and worsening diuretic resistance. NSAIDs are contraindicated in HF-CKD and should be actively withdrawn from the medication list.5 Contrast media: Iodinated contrast for CT or angiography carries the risk of contrast-induced nephropathy (CIN) in HF-CKD. Where possible, use iso-osmolar or low-osmolar contrast at the lowest necessary volume; ensure adequate hydration (volume expansion with normal saline or sodium bicarbonate pre- and post-procedure); hold nephrotoxins (NSAIDs, aminoglycosides) peri-procedurally. Gadolinium: In patients with eGFR <30, gadolinium-based MRI contrast carries the risk of nephrogenic systemic fibrosis (NSF): particularly with older macrocyclic agents. Use macrocyclic gadolinium agents (gadobutrol, gadoteridol) preferentially; avoid linear agents in severe CKD.
HF in pregnancy may arise from pre-existing cardiomyopathy or from peripartum cardiomyopathy (PPCM): a dilated cardiomyopathy of unknown etiology presenting in the last month of pregnancy or within 5 months postpartum in the absence of pre-existing cardiomyopathy.6 PPCM carries a variable prognosis: approximately 50% of patients recover to near-normal LVEF within 6–12 months; the remainder progress to chronic HFrEF or have recurrent disease with subsequent pregnancies. The pharmacological challenge in pregnancy is that several cornerstone HF drugs are teratogenic or harmful to the fetus, requiring careful drug selection and close obstetric-cardiology collaboration.
Diuretics: Loop diuretics (furosemide) may be used in pregnancy for pulmonary edema or severe fluid overload. Caution is required to avoid volume depletion, which can reduce uteroplacental perfusion. Avoid long-term prophylactic diuretic use in pregnancy.6 Beta-blockers: Carvedilol, metoprolol succinate, and bisoprolol may be used in pregnancy for HF if required. Beta-blockers cross the placenta and are associated with intrauterine growth restriction, neonatal bradycardia, and neonatal hypoglycemia; these risks are manageable and are outweighed by the maternal benefit in severe HF. Carvedilol has been used in PPCM. Observe neonates closely for bradycardia and hypoglycemia in the immediate postnatal period.6 Hydralazine: Arterial vasodilator: the preferred afterload-reducing agent in pregnancy (safe record in pregnancy hypertension management). Used in acute hypertensive emergencies and for chronic afterload reduction when RAAS blockade is contraindicated.6 Nitrates (nitroglycerin, isosorbide dinitrate): Safe in pregnancy; used for acute pulmonary edema and venodilation. Avoid at high doses in third trimester due to potential fetal methemoglobinemia with prolonged use. Digoxin: Safe in pregnancy; crosses the placenta but is not teratogenic; widely used for maternal rate control in AF during pregnancy. Monitor levels in the context of volume changes of pregnancy (increased volume of distribution (Vd) and renal clearance may lower levels).6
Heparin: Used for anticoagulation in HF with AF or LV thrombus in pregnancy; does not cross the placenta. Low-molecular-weight heparin (LMWH, e.g., enoxaparin) is preferred; unfractionated heparin is used near term. Warfarin crosses the placenta and is teratogenic in the first trimester (warfarin embryopathy); avoid in T1; may be used with caution in T2–T3 for specific indications (e.g., mechanical heart valves). Bromocriptine: A dopamine agonist that inhibits prolactin secretion; has been evaluated in PPCM based on the hypothesis that oxidative-cleaved prolactin (16-kDa fragment) is cardiotoxic. A pilot RCT (PPCM-Bromocriptine, Germany) and subsequent data suggested improved LVEF recovery in PPCM patients treated with bromocriptine 2.5 mg BID for 4 weeks.9 The 2019 ESC position statement on PPCM gives bromocriptine a Class IIb recommendation for severe PPCM. Note: bromocriptine inhibits breastfeeding; this must be discussed with patients.
ACE inhibitors and ARBs: Contraindicated in all trimesters. In the first trimester: associated with cardiac malformations and neural tube defects. In the second and third trimesters: cause fetal hypotension, oligohydramnios, renal tubular dysgenesis, and neonatal renal failure: a fetal RAAS-blockade syndrome.6 If a pregnant woman with HF is on an ACEi or ARB, it must be stopped immediately upon confirmation of pregnancy and replaced with hydralazine + nitrate. Sacubitril/valsartan (ARNI): Contraindicated in pregnancy, contains valsartan (ARB). Same fetal risks as ARBs. SGLT2 inhibitors: Contraindicated in pregnancy, insufficient safety data; embryotoxic in animal studies at high doses; avoid. Mineralocorticoid receptor antagonists: Spironolactone is contraindicated in pregnancy: it is anti-androgenic and may cause feminization of a male fetus. Eplerenone has insufficient pregnancy data and should also be avoided.6 If decongestion beyond loop diuretics is required, thiazides (with caution) or amiloride (potassium-sparing, non-MRA) may be considered. Statins: Contraindicated in pregnancy due to teratogenicity (fetal limb malformations, CNS defects in animal studies). Stop before conception or immediately upon pregnancy confirmation. Amiodarone: Relative contraindication in pregnancy; may cause neonatal thyroid dysfunction (hypo- and hyperthyroidism), neonatal bradycardia, and intrauterine growth restriction (IUGR). Avoid unless necessary for life-threatening refractory arrhythmia. Ivabradine and vericiguat: Insufficient safety data; avoid in pregnancy.
RAAS blockade (ACEi/ARBs, ARNI) can be safely restarted postpartum; captopril, enalapril, and benazepril have minimal transfer into breast milk and are considered compatible with breastfeeding. MRAs may be restarted postpartum. SGLT2 inhibitors: avoid during breastfeeding (excreted in breast milk in animal studies; insufficient human lactation data). Beta-blockers: metoprolol and carvedilol are compatible with breastfeeding; monitor infant for bradycardia. Women with PPCM who have not achieved LVEF recovery should be strongly counseled against subsequent pregnancy, which carries high risk of PPCM recurrence and maternal death.6
Redfield MM. Heart failure with preserved ejection fraction. N Engl J Med. 2016;375(19):1868–1877
doi:10.1056/NEJMcp1511175Solomon SD, McMurray JJV, Anand IS, et al. Angiotensin-neprilysin inhibition in heart failure with preserved ejection fraction (PARAGON-HF). N Engl J Med. 2019;381(17):1609–1620
doi:10.1056/NEJMoa1908655Heidenreich PA, Bozkurt B, Aguilar D, et al. 2022 AHA/ACC/HFSA guideline for the management of heart failure. J Am Coll Cardiol. 2022;79(17):e263–e421
doi:10.1016/j.jacc.2021.12.012Gheorghiade M, van Veldhuisen DJ, Colucci WS. Contemporary use of digoxin in the management of cardiovascular disorders. Circulation. 2006;113(21):2556–2564
doi:10.1161/CIRCULATIONAHA.105.560110Ronco C, Haapio M, House AA, et al. Cardiorenal syndrome. J Am Coll Cardiol. 2008;52(19):1527–1539
doi:10.1016/j.jacc.2008.07.051Sliwa K, Böhm M. Incidence and prevalence of peripartum cardiomyopathy. Ann Intern Med. 2014;161(12):876–877
doi:10.7326/M14-1416Mehra MR, Uriel N, Naka Y, et al. A fully magnetically levitated left ventricular assist device — final report (MOMENTUM 3). N Engl J Med. 2019;380(17):1618–1627
doi:10.1056/NEJMoa1900486Halliday BP, Wassall R, Lota AS, et al. Withdrawal of pharmacological treatment for heart failure in patients with recovered dilated cardiomyopathy (TRED-HF). Lancet. 2019;393(10166):61–73
doi:10.1016/S0140-6736(18)32216-2Bauersachs J, König T, van der Meer P, et al. Pathophysiology, diagnosis and management of peripartum cardiomyopathy: a position statement from the Heart Failure Association of the European Society of Cardiology Study Group on peripartum cardiomyopathy. Eur J Heart Fail. 2019;21(7):827–843
doi:10.1002/ejhf.1493