ICD Implantation in HFrEF Patients with Improved EF on GDMT
ICD implantation remains indicated for primary prevention in HFrEF patients who have improved their EF with GDMT but still have LVEF ≤35% after at least 3 months of optimal therapy, and counseling about ICD should occur once target or maximally tolerated GDMT doses are achieved. 1
Timing of ICD Decision-Making After GDMT Optimization
The critical 3-month window for ICD evaluation remains appropriate even with contemporary GDMT including ARNI and SGLT2 inhibitors. 1, 2 Current guidelines specify that repeat imaging should occur after 3-6 months of optimal GDMT doses to guide device therapy decisions 1.
Most patients who will experience significant LVEF recovery do so within the first 3 months of intensive GDMT optimization. 2 In a 2025 prospective study, 58% of newly diagnosed HFrEF patients achieved LVEF >35% by 3 months with early intensive GDMT including ARNI and SGLT2i 2.
Patients who fail to recover LVEF >35% by 3 months show limited additional improvement by 1 year (only 18% recovered between 3-12 months), suggesting the 3-month evaluation point remains valid for ICD decision-making 2.
Current Guideline Recommendations for ICD Implantation
Primary Prevention Indications
ICD implantation receives a Class 1 recommendation for patients with LVEF ≤35% who remain symptomatic (NYHA Class II-III) despite ≥3 months of GDMT, with stronger evidence for ischemic cardiomyopathy (Level A) than non-ischemic etiology (Level B). 1
Counseling regarding ICD implantation should occur for all patients with LVEF ≤35% despite ACE inhibitor/ARB/ARNI and beta-blocker therapy for at least 3 months. 1 This counseling must include documentation of discussion about sudden versus non-sudden death risk, ICD efficacy, safety, potential complications, and the option to deactivate the device 1.
Patients Who Should NOT Receive ICD
ICD implantation is of uncertain benefit (Class IIb) in patients with high risk of non-sudden death, including those with frequent hospitalizations, advanced frailty, or severe comorbidities limiting survival. 1
ICD is not indicated when comorbidities and/or frailty limit survival with good functional capacity to <1 year. 1
Management of Patients with Improved EF (HFimpEF)
Continuation of GDMT is Critical
Patients with previous HFrEF whose EF improves to >40% should continue their complete HFrEF treatment regimen indefinitely. 3, 4 This represents a new category termed "HFimpEF" (heart failure with improved ejection fraction) 1.
Discontinuation of HFrEF medications after EF improvement may lead to clinical deterioration. 3, 4 The improved EF reflects the therapeutic effect of GDMT, not resolution of the underlying cardiomyopathy 3.
ICD Decision-Making in Improved EF Patients
For patients whose EF improves above 35% during the initial 3-6 month GDMT optimization period (before ICD implantation), ICD is no longer indicated for primary prevention. 1 These patients should continue GDMT and undergo periodic reassessment 1.
The situation differs for patients who already have an ICD and subsequently improve their EF >35%. 5 While limited data exist, the risk of sudden cardiac death in these patients remains uncertain, and decisions about generator replacement at battery depletion require individualized assessment 5.
Impact of Contemporary GDMT on Sudden Cardiac Death Risk
Evidence for Reduced SCD Risk with Modern Therapy
Contemporary quadruple GDMT (ARNI, beta-blocker, MRA, SGLT2i) reduces all-cause mortality by approximately 73% over 2 years and has favorable effects on sudden cardiac death reduction. 3, 6 Each component contributes to SCD prevention: beta-blockers and MRAs have established anti-arrhythmic effects, ARNI mitigates adverse ventricular remodeling, and SGLT2 inhibitors show beneficial effects in reducing serious ventricular arrhythmias 6.
The landmark ICD trials (MADIT-II, SCD-HeFT) that established current guidelines preceded the era of ARNI and SGLT2 inhibitors. 7, 8 This has led some experts to question whether ICD continues to provide the same magnitude of benefit at current LVEF cutoffs for patients on contemporary GDMT 7, 6.
Current Guideline Position Remains Unchanged
Despite advances in medical therapy, current 2022 ACC/AHA/HFSA guidelines maintain the LVEF ≤35% threshold and 3-month GDMT optimization period for ICD consideration. 1 The guidelines acknowledge that persistently reduced LVEF ≤35% despite GDMT for ≥3 months warrants referral for device therapy consideration 1.
Ongoing trials are evaluating whether contemporary GDMT alters ICD benefit, but until definitive evidence emerges, current recommendations should be followed. 8
Practical Algorithm for ICD Decision-Making
Step 1: Optimize GDMT (Months 0-3)
Initiate all four foundational GDMT classes simultaneously at low doses: ARNI (preferred over ACEi/ARB), beta-blocker (carvedilol, metoprolol succinate, or bisoprolol), MRA (spironolactone or eplerenone), and SGLT2i (dapagliflozin or empagliflozin). 3, 4
Uptitrate every 1-2 weeks to target or maximally tolerated doses, checking blood pressure, renal function, and electrolytes after each increment. 3, 4
Step 2: Reassess LVEF at 3 Months
- Perform repeat echocardiography after 3 months of optimal GDMT doses (or maximally tolerated doses if target doses cannot be achieved). 1, 2
Step 3: Apply ICD Criteria
If LVEF remains ≤35%:
- Counsel patient about ICD for primary prevention if NYHA Class II-III symptoms persist and life expectancy >1 year without severe comorbidities. 1
- Proceed with ICD implantation (Class 1 recommendation) unless patient declines after informed discussion or has contraindications. 1
If LVEF improved to >35%:
- ICD is not indicated for primary prevention. 1
- Continue all GDMT medications indefinitely. 3, 4
- Consider repeat imaging at 6-12 months or with clinical status changes. 1
Step 4: Special Considerations
For patients with anticipated need for frequent ventricular pacing (>40%), consider CRT-D even if LVEF improves, as pacing-induced cardiomyopathy risk remains. 1
For non-ischemic cardiomyopathy, the evidence for ICD benefit is less robust (Class IIa, Level B-R in some guidelines), warranting more detailed shared decision-making discussions. 1, 6
Common Pitfalls to Avoid
Do not delay ICD evaluation beyond 3-6 months hoping for further LVEF improvement in patients who remain ≤35% at 3 months. 2 The data show minimal additional recovery occurs after this point 2.
Do not discontinue GDMT in patients whose EF improves above 35%. 3, 4 The improved EF is a treatment effect, not disease resolution 3.
Do not implant ICD before completing 3 months of optimal GDMT (except in specific scenarios like prior cardiac arrest or sustained VT) 1. Premature implantation may subject patients to unnecessary device therapy if their EF would have recovered 2.
Do not use asymptomatic low blood pressure (systolic 80-100 mmHg with adequate perfusion) as a reason to avoid GDMT uptitration or delay ICD evaluation. 1, 3 Patients with adequate perfusion tolerate these blood pressures well 3.