End-stage heart failure (HF) accounts for 1% to 10% of all HF cases. It is frequently associated with coexistent atrial fibrillation due in part to structural changes related to altered hemodynamics, increased wall stress, and neurohormonal activation. The treatment remains challenging because of frequent atrial remodeling, comorbidities, and high recurrence rates. Studies have found no benefit of medical rhythm control compared with rate control. However, recent data suggest that catheter ablation might be associated with a reduction in HF symptomatology and all-cause mortality. Alternative treatment options including AV nodal ablation are available for refractory cases, but should not delay urgent heart transplantation.
Key points
- •
Up to 50% of patients with end-stage heart failure have concomitant atrial fibrillation.
- •
End-stage heart failure was previously thought to indicate that a benefit from catheter ablation for atrial fibrillation is unlikely.
- •
Recent data suggest that catheter ablation might be associated with a reduction in heart failure symptoms and all-cause mortality.
Background
Definition and Epidemiology of End-Stage Heart Failure
End-stage heart failure (HF), considered to be synonymous with ‘advanced “ or ‘refractory” HF, indicates the persistence of severe symptoms despite optimal medical therapy. It includes episodes or cardiac decompensation requiring hospital admission and is associated with a severely impaired exercise capacity such as a 6 minute walk test distance (6MWTD) less than 300 m or a peak V o 2 (pVO2) less than 12 to 14 mL/kg/min. While a severely reduced LVEF is common, end-stage HF may also occur in patients with cardiac dysfunction related to valvular abnormalities, congenital abnormalities, right ventricular failure, or severe diastolic dysfunction. The International Society for Heart Lung Transplantation has published criteria to guide heart transplantation listing for patients with the most advanced disease, including an extensive evaluation, clinical risk scores, and an assessment of comorbidities.
End-stage HF accounts for 1% to 10% of all heart failure cases, with a higher prevalence in older patients. , The prevalence of end-stage HF is expected to increase due to demographic changes and a rising number of overall HF cases. However, the prognosis remains poor, with 1-year mortality rates up to 75%.
Atrial Fibrillation in End-Stage Heart Failure
Overall, 10% to 50% of patients with HF have concomitant AF, , due to a higher incidence of AF among patients with HF and a higher incidence of HF among patients with AF ( Fig. 1 ). Data from the Framingham Heart Study participants show that HF is linked to the development of incident AF (hazard ratio 2.18, 95% Confidence Interval 1.26–3.76), and prevalent AF is associated with incident HF, particularly heart failure with preserved ejection fraction (HFpEF). Notably, the prevalence of AF in patients with HF increases as the severity of HF increases, approaching 50% in patients with New York Heart Association (NYHA) functional class IV. ,
Concomitant AF is a predictor of worse outcomes. Whether AF is itself contributing to worse outcomes or whether it is simply a marker of advanced disease was initially controversial. Earlier studies in patients with advanced HF did not identify AF as an independent predictor after adjusting for age, left ventricular ejection fraction (LVEF), and comorbidities. Other smaller studies also failed to demonstrate an independent association between AF and worse outcomes after controlling for additional covariates. However, analyses of cohorts such as the Framingham Study indicate that the subsequent development of AF in patients with HF was associated with a worse overall prognosis. Those temporal relations provide strong support for the contribution of AF to adverse outcomes in patients with HF.
The coexistence of AF and HF is associated with an increased risk for HF hospitalizations and all-cause mortality. , In a study of 390 patients with severe HF (NYHA functional class III or IV, mean LVEF 19%), AF was associated with an increased risk of overall and sudden death. Notably, there was no difference between paroxysmal and chronic AF. However, AF was only associated with mortality in patients with a pulmonary capillary wedge pressure (PCWP) less than 16 mm Hg, possibly indicating a diminishing impact on patient outcomes in those with the most advanced HF. Supraventricular tachycardias have also been linked to an increased rate of hospitalization for HF, increased risk of stroke, and total mortality in the DIG trial, which included 7788 with HF (NYHA functional class III or IV, 31.4%).
Discussion
Pathophysiology
Electrophysiological studies showed that a reduced LVEF and symptomatic HF are associated with regions of low-voltage amplitude and scarring in both atria, impaired atrial conduction, and increased inducibility and duration of atrial fibrillation. On a cellular level, these changes may be related to an increase in interstitial fibrosis and subsequent alteration of local conduction. , In addition, cellular hypertrophy and degeneration have been observed. These changes may lead to a left atrial (LA) myopathy, which manifests via eccentric LA remodeling and a reduced LA ejection fraction and provides the substrate for the development of AF. The underlying mechanisms for the development of structural changes in the form of atrial fibrosis appear to be related to a decrease in cardiac output, increased vascular resistance, increased LA wall stress, and neurohormonal changes including elevated levels of norepinephrine and endothelin and RAAS overactivation.
Experimental studies have also suggested calcium-handling abnormalities underlying atrial arrhythmogenesis in HF. HF was associated with an increased LA diastolic intracellular calcium concentration and sarcoplasmic reticulum calcium load, which was associated with spontaneous calcium transient events and triggered ectopic activity. Other studies have supported the role of enhanced sarcoplasmic reticulum diastolic calcium leak in causing AF-promoting delayed afterdeloplarization. Together, these findings suggest abnormal impulse formation and propagation underlying the development of AF in HF.
AF may in turn contribute to HF exacerbations mediated by hemodynamic changes associated with the loss of atrial contraction, an irregular ventricular rhythm, and chronotropic dysregulation. The fall in cardiac output and an irregular cardiac rhythm caused by AF lead to neurohormonal activation, which is associated with an increase in afterload and further compounding of the changes observed in HF. In addition, tachycardia and an irregular ventricular rhythm might contribute to a reversible LV systolic dysfunction. A degree of tachycardia-mediated cardiomyopathy (TMC) might be seen in a significant number of patients with a dilated cardiomyopathy and new AF, but the exact prevalence remains unknown. The hemodynamic benefits and the potential improvement in LV function provide the theoretic basis for a rhythm control strategy even in patients with end-stage HF.
Guideline Recommendations for the Management of Atrial Fibrillation in Advanced Heart Failure
While the 2023 ACC/AHA/ACCP/HRS AF guidelines recommend catheter ablation in patients with AF and heart failure with reduced ejection fraction (HFrEF) to improve symptoms, quality of life, ventricular function, and cardiovascular function (class of recommendation I, level of evidence A), no specific recommendations are included for patients with end-stage HF. However, the guidelines provide several criteria to evaluate if patients are appropriate for rhythm control with catheter ablation. Advanced HF, significant ventricular scar, severe atrial myopathy, long-standing persistent AF, advanced age, and multiple comorbidities indicate that a benefit from catheter ablation is unlikely. The most recent 2020 ESC atrial fibrillation guidelines provide no specific recommendations for patients with AF and end-stage HF.
Medical Therapies in End-Stage Heart Failure – Rate Versus Rhythm control
Early studies conducted more than two decades ago did not find a difference in outcomes between medical rate and rhythm control strategies in patients with AF. However, only a small number of patients enrolled in these trials had LV dysfunction and none had end-stage HF. The largest randomized controlled trial comparing medical rate and rhythm control in patients with symptomatic HF included 1376 patients with an LVEF ≤35%. Overall, the mean LVEF was 27% and 32% of patients had NYHA functional class III or IV at baseline. Most patients had been hospitalized for HF within the prior 6 months. Notably, patients with long-standing persistent AF were excluded, as were those with expected cardiac transplantation within 6 months or a life expectancy of less than 1 year. There was no difference in the primary outcome of time to cardiovascular death after a mean of 37 months in the overall study cohort, in patients with an LVEF ≤25%, or those with NYHA class III or IV ( Fig. 2 ). Similarly, there was no difference in all-cause mortality, worsening HF, or stroke. Amiodarone was used in 82% of patients in the rhythm control arm, while 75% of patients in the rate control arm were treated with digoxin. The role of digoxin as a rate-control agent in patients with AF and HF, particularly those with end-stage HF, has been of special interest.
The current ACC/AHA/ACCP/HRS guidelines still recommend digoxin for rate control as an alternative to beta blockers or nondihydropyridine calcium channel blockers (in patients with an LVEF >40%) with a class IIa indication. However, concerns have persisted regarding an increase in adverse events associated with the use of digoxin. , Meta-analyses of observational studies have found an increased risk of all-cause mortality and cardiovascular mortality in patients with AF, regardless of the presence of HF. , Conversely, meta-analyses of randomized trials found a neutral effect on all-cause mortality. , Recently, the RATE-AF trial compared digoxin to bisoprolol in patients with AF and coexisting HF. Overall, 19% had a reduced LVEF and 38% had NYHA functional class III or IV. While there was no significant difference in quality of life at 6 months, several quality of life scores as well as the NYHA functional class at 12 months favored digoxin. Overall, digoxin remains an important option for rate control, especially in patients with AF and end-stage HF if other medical therapies have failed.
Atrial Fibrillation Ablation
Several trials have investigated the role of catheter ablation in patients with AF and concomitant HF ( Table 1 ). Overall, the results have suggested a reduction in all-cause mortality, HF hospitalizations, LVEF improvement, and quality of life. However, few studies have included patients with end-stage HF. Moreover, CASTLE-AF, RAFT-AF, AATAC, and CAMTAF excluded patients with a limited life expectancy or if a cardiac intervention or heart transplant was likely within 6 to 12 months after enrollment. MacDonald et al. included patients with relatively advanced HF, including 90% with NYHA class III, and showed no difference in the primary outcome (change in LVEF on cardiac magnetic resonance [CMR]). However, the study only included a total of 41 patients and excluded patients if a cardiac transplant was likely within 6 months. The AMICA trial also enrolled 140 patients with more advanced HF, with a mean LVEF of 26% and 61% of patients with NYHA functional class III. Additionally, 44% of those patients were treated with a cardiac resynchronization therapy with a defibrillator (CRT-D). The trial failed to meet its primary endpoint of change in LVEF at 1 year due to a similar improvement in the medical therapy arm. It also found no difference in 6-min walk distance (6MWD) and quality of life indicators. Notably, catheter ablation appeared to be able to effectively reduce AF burden even in this population of patients with advanced HF as 72% of patients had an AF burden of ≤5% after ablation.
| N | Persistent AF | LVEF | NYHA III | NYHA IV | 6MWD | Peak V o 2 | MLWHF | Intervention | Primary Outcome | Notable Secondary Outcomes | Comments | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| PABA-CHF (Khan et al, 2008) | 81 | 48% | 28±10% | Unknown | – | 275± m | – | 89±16 | PVI ± CFAE, linear lesions based on center preference vs AV nodal ablation with biventricular pacing | Composite of MLWHF and LVEF at 6 mo ( improved in PVI arm ) | Greater improvement in LVEF, 6MWD, and MLWHF score in non-paroxysmal groups | Exclusion if expected cardiac transplant within 12 mo or life expectancy <2 y |
| MacDonald et al, Heart 2011 | 41 | 100% | Mean: 17±9% (radionuclide scan) | 90% | – | 334± 172m | – | 57±30 | PVI, linear lesions if AF persisted after PVI vs rate control | Change in CMR LVEF ( no difference ) | Change in radionuclide LVEF (increased in PVI group), otherwise no difference | Exclusion if expected cardiac transplant within 6 mo |
| ARC-HF (Jones et al, 2013) | 52 | 100% | 24±11% (radionuclide scan) | 48% | – | 414± 134m | 17±7 | 46±34 | PVI, linear lesions if AF persisted after PVI vs rate control | 12-mo change in peak V o 2 ( increased in ablation arm ) | Improved QoL and BNP in ablation arm | Exclusion if urgent transplant listing |
| CAMTAF (Hunter et al, 2014) | 55 | 100% | 33±14% | 58% | – | – | – | – | PVI + CFAE ablation ± linear lesions if AF persisted after initial ablation vs rate control | LVEF change at 6 mo ( increased in ablation arm ) | Catheter ablation was associated with improved peak V o 2 and QoL | Exclusion if the expectation of device therapy or cardiac surgery within 12 mo |
| AATAC (Di Biase et al, 2016) | 203 | 100% | 29±9% | Unknown | Unknown | 349± 171m | – | 51±36 | PVI + posterior wall isolation ± SVC isolation, CFAE ablation, linear lesions vs amiodarone | Freedom from AF, atrial flutter, or atrial tachycardia >30 s at 24 mo ( improved in PVI group ) | All-cause mortality and unplanned hospitalization (both lower in ablation arm) | All patients had an ICD or CRT-D; exclusion if life expectancy ≤2 y |
| CAMERA-MRI (Prabhu et al, 2017) | 68 | 100% | 33±12% | 50% | Unknown | 490± 198m | – | – | PVI + posterior wall isolation ± AADs vs rate control | Change in LVEF at 6 mo ( increased in ablation arm ), response more pronounced in patients without LGE | Average AF burden 1.6 ± 5.0% at 6 mo in ablation arm (100% in rate control arm) | All patients had idiopathic cardiomyopathy; 48% of patients had an LVEF <35% |
| CASTLE-AF (Marrouche et al, 2018) | 363 | 70% | 32% (Q1 26%, Q3 38%) | 28% | 1% | – | – | – | PVI + additional lesions to restore sinus rhythm vs medical treatment (efforts to maintain sinus rhythm recommended) | All-cause death or hospitalization for HF ( reduced in ablation arm ), no difference in patients with LVEF <25% or NYHA III, but few events | Individual outcomes (all-cause death, cardiovascular death, and hospitalization for HF) reduced in ablation arm | 72% of patients had an ICD, 28% had a CRT-D; exclusion if candidate for heart transplant or cardiovascular intervention |
| AMICA (Kuck et al, 2019) | 140 | 100% | 26±13% | 61% | – | – | – | – | PVI ± additional lesions based on operator preference vs medical therapy (rate or rhythm control) | Change in LVEF at 1 y ( no difference ) | No difference in 6MWD, QoL, BNP; AF burden ≤5% in 72% of patients in the ablation arm | 56% of patients had an ICD, 44% had a CRT-D; |
| CABANA (Packer et al, 2019, Packer et al, , 2021) | 2204, HF: 778 | 55% among patients with HF | 55% (Q1 50%, Q3 61%) | 24% | 0.3% | – | – | – | PVI ± additional lesions based on operator preference vs medical therapy (rate control recommended as initial strategy) | Composite of all-cause mortality, disabling stroke, serious bleeding, or cardiac arrest ( reduced in ablation arm in patients with HF ) | In HF patients, reduced all-cause mortality and AF recurrence with catheter ablation. Improved QoL in ablation arm. | 7.9% of HF patients had an LVEF <35% |
| RAFT-AF (Parkash et al, 2022) | 411 | 93% | 58% had LVEF ≤45% (mean LVEF 30%) | 36% of patient with an LVEF ≤45% | – | 350± 145m among patients with LVEF ≤45% | – | – | PVI + additional lesions in patients with persistent AF vs rate control | Composite of all-cause mortality and all HF events ( not significantly different in the overall analysis or among patients with LVEF ≤45%) | LVEF, 6MWD, QoL (all improved in ablation arm) | 18% of patients with an LVEF ≤45% had a CRT-D; exclusion if life expectancy <1 y |
| CASTLE-HTx (Sohns et al, 2023) | 194 | 70% | 27±8% | 55% | 14% | 304± 95m | – | – | PVI ± additional lesions based on operator preference vs medical therapy with recommendation to maintain sinus rhythm | Composite of all-cause mortality, LVAD implantation, and urgent heart transplantation ( reduced in ablation arm ) | Reduced all-cause mortality, cardiovascular mortality, and LVAD implantation in ablation arm | All patients were referred for the assessment of eligibility for heart transplant; all patients had an implantable cardiac device with arrhythmia detection (38% had a CRT-D); exclusion if listed as “high urgent” for heart transplantation |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
