Targeting the Substrate for Atrial Fibrillation

Abstract

The identification of the pulmonary veins as a trigger source for atrial fibrillation (AF) has established pulmonary vein isolation (PVI) as a key target for AF ablation. However, PVI alone does not prevent recurrent AF in many patients, and numerous additional ablation strategies have failed to improve on PVI outcomes. This therapeutic limitation may be due, in part, to a failure to identify and intervene specifically on the pro-fibrillatory substrate within the atria and pulmonary veins. In this review paper, we highlight several emerging approaches with clinical potential that target atrial cardiomyopathy—the underlying anatomic, electrical, and/or autonomic disease affecting the atrium—in various stages of practice and investigation. In particular, we consider the evolving roles of risk factor modification, targeting of epicardial adipose tissue, tissue fibrosis, oxidative stress, and the inflammasome, along with aggressive early anti-AF therapy in AF management. Attention to combatting substrate development promises to improve outcomes in AF.

Identification of the pulmonary veins as a trigger source for atrial fibrillation (AF)¹ established pulmonary vein isolation (PVI) as a key ablation target. Yet, PVI alone does not prevent recurrent AF in many patients, particularly those with persistent AF. Various approaches have been tested to improve ablation outcomes, generally with the conceptual framework that additional ablation is the solution. These approaches can be classified as empiric, electrogram guided, or substrate focused. The classic empiric approach of delivering linear lesions in the atrium provided no advantage over PVI.² Targeting complex fractionated electrograms did not improve outcomes.² Mapping approaches, such as focal impulse and rotor modulation have not provided clear benefit.³ Recently, targeting fibrosis regions detected by cardiac magnetic resonance imaging did not improve outcome but did increase adverse events.⁴ Empiric isolation of the posterior left atrial (LA) wall also provided no incremental benefit to PVI alone.⁵ Furthermore, performing extensive LA ablation to reduce the volume of conducting atrial tissue could be effective in preventing AF, but could also cause extensive fibrosis impairing LA function.⁶ Given the plethora of well-intended but failed approaches to improve outcomes by “extending the lesion set,” it is critical to contemplate whether there are promising alternative pathophysiologic-based approaches that target the underlying atrial cardiomyopathy (ACM) that is responsible for AF (Central Illustration). There are known clinical causative factors and mechanistic mediators for which targeted interventions have been studied (Figure 1) but not yet fully integrated into a comprehensive therapeutic AF treatment strategy that may include individualized ACM therapies in conjunction with ablation. Our inability to identify patient-specific AF substrates to target for treatment presents a therapeutic barrier that should be the focus of further research efforts. Here, we highlight several promising pathophysiologic-based approaches that target ACM—the underlying anatomic, electrical, and/or autonomic dysfunction affecting the atrium—that may have potential clinical value.

CENTRAL ILLUSTRATION.

Clinical Factors and Mechanistic Mediators Underlying Atrial Cardiomyopathy and Atrial Fibrillation Pathogenesis

Clinical causative factors (green) and mechanistic mediators (orange) underlying the pathogenesis of atrial cardiomyopathy (ACM) and atrial fibrillation (AF). Upstream opportunities for control of risk conditions and targeted therapeutics to impact the mechanisms leading directly to ACM may improve AF outcomes.

FIGURE 1.

Target Therapies for Upstream Clinical Factors and Mechanistic Mediators

Many of the upstream clinical causative factors and underlying mechanistic mediators that generate atrial cardiomyopathy ACM have associated targeted therapies (blue boxes), but for only a few of these interventions is there current evidence of benefit in clinical studies. *Evidence of benefit in clinical studies. AF ¼ atrial fibrillation; BP¼blood pressure; CAD¼coronary artery disease; Casp-1 ¼caspase 1; CKD¼chronic kidney disease; COPD¼chronic obstructive pulmonary disease; CPAP¼continuous positive airway pressure; GDMT ¼ guideline-directed medical therapy; GLP1 ¼ glucagon-like peptide 1; IL ¼ interleukin; NLRP3 ¼ NACHT, LRR, and PYD domain-containing protein-3; NOX2 ¼ nicotinamide adenine dinucleotide phosphate oxidase enzyme 2; OSA ¼ obstructive sleep apnea; RAS ¼ renin-angiotensin system; ROS ¼ reactive oxygen species; SGLT2 ¼ sodium-glucose cotransporter 2; TGF ¼ transforming growth factor; TNF ¼ tumor necrosis factor; XO ¼ xanthine oxidase.

RISK FACTOR MODIFICATION: TARGETING OBESITY AND OTHER ADVERSE LIFESTYLE FACTORS

The AF guidelines⁷ include a Class I indication for “weight loss combined with risk factor modification,” largely based on studies identifying mechanistic links between obesity and AF and the beneficial effects of weight loss. Epidemiologic studies demonstrate association between body mass index (BMI) and AF risk. Excessive body weight is more strongly associated with the development of persistent/permanent AF than paroxysmal AF.⁸ Higher BMI is associated with increased risk of recurrence after catheter ablation.⁹ Obesity is associated with atrial structural abnormalities. However, the exact mechanism delineating how obesity contributes to AF is still unclear, although multiple electrical, anatomical, metabolic, and hemodynamic factors are involved. Targeting obesity and its comorbidities presents an additional therapeutic target for AF. In a retrospective study of 220 morbidly obese patients with AF (BMI ≥40 kg/m²) undergoing bariatric surgery, AF “regression” (persistent to paroxysmal, or resolution) occurred in 71% after gastric bypass, 56% after gastrectomy, and 50% after gastric banding.¹⁰ In ARREST-AF (Aggressive Risk Factor Reduction Study for Atrial Fibrillation),¹¹ patients undergoing AF ablation with BMI ≥27 kg/m² and ≥1 cardiac risk factor were offered either risk factor management and ablation (n = 61) or ablation alone (n = 88). The former group achieved larger reduction in weight (13.2 ± 5.4 kg vs 1.5 ± 5.1 kg, P = 0.002), as well as lower blood pressure, HbA1C, and increased continuous positive airway pressure usage. At average follow-up of ~42 months, single (32.9% vs 9.7%, P < 0.001) and multiple (87.% vs 17.8%, P < 0.001) procedure freedom from arrhythmia improved in the risk factor management group. In LEGACY¹² (Long-Term Effect of Goal-directed weight management on Atrial Fibrillation Cohort: A 5-Year follow-up study), in those who achieved ≥10% weight loss, there was a 6-fold (95% CI: 3.4–10.3; P < 0.001) increase in arrhythmia-free survival compared with patients with <3% and 3% to 9% weight loss. REVERSE-AF¹³ (PREVEntion and regReSsive Effect of weight-loss and risk factor modification on Atrial Fibrillation) explored change in AF type over time by weight loss group in LEGACY. In those with ≥10% weight loss, 3% progressed to persistent AF and 88% “regressed” from persistent to paroxysmal AF, compared with 41% and 26% in those with <3% weight loss and 32% and 49% in those with 3% to 9% weight loss, respectively. Some studies have shown no benefit to weight loss.^14,15

Some medications that promote weight loss have been associated with improvement in AF. In a meta-analysis, sodium-glucose cotransporter 2 inhibitor therapy was associated with a 19% reduction in the odds of incident atrial arrhythmias compared with control.¹⁶ Glucagon-like peptide-1 receptor (GLP-1R) analogs, such as liraglutide, semaglutide, and dulaglutide, have been shown to cause significant weight loss and reduce major adverse cardiovascular events. The LEAF (Liraglutide Effects in Atrial Fibrillation; NCT03856632) trial is a randomized controlled trial that compares risk factor modification or risk factor modification plus liraglutide in patients undergoing catheter ablation and showed improved outcomes with liraglutide.¹⁷ It is important to note that these medications have pleiotropic effects that may contribute to amelioration of AF and multiple mechanisms may account for improved outcomes with weight loss. In a sheep model, sustained obesity was associated with increased LA pressure, shorter refractory periods, reduced conduction velocity, perivascular inflammation, increased epicardial adipose tissue (EAT) and interstitial fibrosis, increased atrial transforming growth factor (TGF)-β1, and diminished connexin-43; these changes were largely reversed with 30% weight loss.¹⁸ There is clearly a role for obesity in generating ACM that forms the AF substrate, as well as for the ability of weight loss to reverse this substrate. The strength of association between visceral adiposity and AF is stronger for EAT than for overall adiposity, suggesting that this might be a more proximate therapeutic target.

Other components of the risk factor management approach including blood pressure control, exercise, glycemic control, treating obstructive sleep apnea, and smoking cessation, while clinically desirable, when implemented individually lack strong evidence for improving AF outcomes. Yet, trials of renal denervation in hypertensive patients with AF¹⁹ and alcohol cessation improve AF outcomes.²⁰

TARGETING EAT

EAT is unique for its unobstructed contiguity with the heart and its transcriptome and secretome, which differ from that of other fat depots. EAT is not evenly distributed throughout the heart.²¹ Each regional EAT location has peculiar physiological and pathological properties. EAT is as an independent predictor of AF development and recurrence after ablation.²² EAT thickness/volume are greater in patients with chronic, persistent AF than in those with paroxysmal AF, independent of obesity and other traditional risk factors. EAT can contribute to AF through a complexity of mechanisms: genetic, inflammation, fibrosis, fatty infiltration, neural, and electrical modulation of the atria. Peri-atrial EAT has a unique transcriptome and secretome with potential arrhythmogenic properties.²³ The lack of muscle fascia separating peri-atrial EAT from the underlying myocardium and a shared microcirculation allow for bidirectional crosstalk. Interestingly, the arrhythmogenic role of EAT could start with its embryogenesis. Atrial EAT adipocytes originate from differentiation of epicardial progenitor cells and from the secretome of atrial cardiomyocytes.²⁴ The adipogenic potential of atrial cells, also through secretion of atrial natriuretic peptide, is greater in patients with AF than in those without. Under mechanical and hemodynamic stress, the epicardium is reactivated and contributes to the expansion and fibro-fatty infiltration of the peri-atrial EAT.²⁵ Peri-atrial EAT is enriched in genes encoding proteins involved in inflammation and fibrosis. Pro-inflammatory adipokines such as interleukins and tumor necrosis factor (TNF), and profibrotic factors, such as matrix metalloproteinases, connective tissue growth factor, TGF-β1, TGF-β2, and activin A, can diffuse from EAT into the adjacent atrial myocardium and promote AF.^26,27 EAT-derived extracellular vesicles collected from patients with AF contain profibrotic cytokines and microRNAs.²⁸ Peri-atrial EAT accumulation can cause fibrosis, thereby slowing conduction and creating zones of conduction block.²⁹ EAT can serve as a physiological source of free fatty acids for the contiguous atrium.²¹ Under pathological circumstances, free fatty acids can be transported from EAT to the atrial myocardium. Free fatty acid infiltration can separate cardiomyocytes, resulting in conduction slowing, loss of side-to-side cell connections, and myocardial disorganization that leads to conduction delay/block and re-entry.^30,31 Activation of EAT ganglia can shorten action potential duration and increase calcium transient amplitude contributing to the initiation and maintenance of AF.³¹

EAT is a modifiable therapeutic target. EAT can be measured either with echocardiography or cardiac computed tomography.²¹ EAT thickness has shown to shrink in response to pharmacological agents. The presence of GLP-1R within the EAT adipocyte suggests a direct effect of these agents on EAT.³² Activation of EAT GLP-1R may induce free fatty acid oxidation and utilization that would reduce ectopic atrial fat and arrhythmogenicity.³³ The clinical measurability of EAT and its responsiveness to drugs (such as GLP-1R analogs), provides a novel therapeutic approach for AF treatment.

TARGETING ATRIAL FIBROSIS

Atrial fibrosis was first linked to ACM and AF in an experimental model of tachycardiomyopathy, and is a central feature in the ACM phenotype. In a recent consensus document, fibroblast-dependent and mixed cardiomyocyte-fibroblast–dependent ACMs comprise one-half of known ACM histologic types.³⁴ Contributory cellular mechanisms to AF-related fibrosis are myriad: rapid atrial firing contributes to a host of autocrine and paracrine signaling cascades that lead to the secretion of pro-fibrotic factors such as TGF-β1 and TNF-α, remodeling of K⁺ channels and transient receptor potential channels that control fibroblast function, and disorders of intracellular Ca²⁺ release leading to mitochondrial dysfunction.^35,36 Activation of myofibroblasts feeds back to further perpetuate AF through tissue-level changes in myocardial conduction.³⁷ Extracardiac signals such as angiotensin-II and oxidation promote both AF and cardiac fibrosis. Thus, atrial fibrosis appears to promote a positive feedback loop in promoting AF. Selective inhibition of fibrosis may ameliorate AF.

To date, most data supporting an antifibrosis strategy in ACM come from animal models and clinical trials evaluating U.S. Food and Drug Administration–approved drugs for treatment of heart failure. These drugs include angiotensin-converting enzyme inhibitors (ACEIs), angiotensin receptor blockers, beta-blockers, and mineralocorticoid-receptor antagonists. ACEIs are by far the most studied drug class for prevention of atrial fibrosis. In experimental models, dogs with ventricular tachypacing-induced cardiomyopathy experienced interstitial fibrosis, atrial conduction slowing, and heart failure, whereas similarly tachypaced enalapril-treated dogs had significant attenuation of atrial fibrosis, atrial remodeling, and AF.³⁸ Similarly, in 24 rapid atrial-paced dogs, atrial effective refractory period (AERP, a sign of profibrillatory atrial remodeling) shortening was attenuated by captopril.³⁹ Clinical trials of ACEI in heart failure show significant reduction in AF.^40–42 Two meta-analyses of ACEI and angiotensin-receptor-blocker therapies showed a relative risk reduction between 28%⁴³ and 35%.⁴⁴ Similarly, spironolactone, a mineralocorticoid-receptor antagonist, has been shown to limit both interstitial fibrosis and AF vulnerability in spontaneously hypertensive rats⁴⁵ and in a post-myocardial infarction rat model.⁴⁶ In humans, eplerenone reduced AF incidence in patients with systolic heart failure by 42%.⁴⁷

Beta-blockers are hypothesized to limit atrial remodeling and AF primarily by reversing neurohormonal effects of sympathoexcitation; however, whether beta-blockers can limit progression of fibrosis and AF onset is highly controversial. Metoprolol has been shown in animal models to reduce atrial fibrosis. In a canine model of obstructive sleep apnea, atrial structural remodeling and sympathetic innervation were sharply inhibited by continuous infusion of metoprolol.⁴⁸ In contrast, results in humans have been mixed. In an observational study of patients undergoing cardiac surgery, surgical biopsies of LA appendages were obtained; cultured myocytes were then exposed to metoprolol or isoproterenol for 4 days. Metoprolol-treated cardiomyocytes had a reduced fibrotic-like CD90+ subpopulation, with reduced expression of collagen-I and enhanced regenerative potential with upregulation of miR-1, miR-133, and miR-101.⁴⁹ In patients receiving beta-blockers for hypertension, beta-blockers were independently associated with reduced LA longitudinal strain, reduced LA expansion index, and reduced LA emptying fraction.⁵⁰ There is insufficient evidence to support beta-blocker use for prevention of atrial fibrosis and AF.

New experimental antifibrosis strategies include TGF-β1 inhibitors (pirfenidone, nintedanib), TNF-α inhibitors (infliximab), and antioxidants (MitoTEMPO). Pirfenidone reduced atrial fibrosis and AF risk in a ventricular tachypaced canine model.⁵¹ In addition to inhibiting TGF-β1, pirfenidone also reduces expression of profibrotic factors TNF-α, interleukin (IL)-4, and IL-13. Nintedanib may have a role in reducing atrial fibrosis; however, there is a possible drug interaction with direct oral anticoagulants that must be further investigated.⁵² Infliximab, in a small cohort of patients with rheumatoid arthritis, normalized atrial remodeling associated with the disease.⁵³ In a mouse model of obesity-induced AF, mitochondrial-targeted antioxidants such as MitoTEMPO, reduce atrial fibrosis and profibrillatory remodeling.⁵⁴

TARGETING OXIDATIVE STRESS

Reactive oxygen species (ROS) are unstable free radicals produced by various sources within cardiomyocytes including mitochondria, nicotinamide adenine dinucleotide phosphate oxidase enzyme (NOX), peroxisomes, and uncoupled nitric oxide synthase. Imbalance between ROS production and antioxidant capacity leads to oxidative stress, a condition implicated in arrhythmogenesis.⁵⁵ Early work demonstrated increased ROS is associated with shortened AERP in pacing-induced AF dog models, which can be mitigated by the antioxidant vitamin C.⁵⁵ The causal relationship between ROS and AF is complex. First, increased ROS or reactive nitroxide species can directly modulate the electrophysiological properties of several ion channels and Ca²⁺-handling proteins via oxidation or nitrosylation.⁵⁶ ROS/reactive nitroxide species can also alter the activities of key protein kinases (eg, protein kinase A, Ca²⁺/calmodulin-dependent protein kinase II) and phosphatases, which in turn modulate the functions of ion channels and Ca²⁺-handling proteins. Second, ROS enhances c-Src tyrosine kinase activity, reducing gap junction protein connexin-43 levels and impairing conduction.⁵⁷ Third, ROS can initiate inflammation and fibrotic remodeling, subsequently promote gap junction remodeling and impair atrial conduction. Moreover, ROS can damage DNA in nuclei and mitochondria, leading to electrophysiological impairment.⁵⁸ During AF, elevated cytosolic Ca²⁺ level in atrial myocytes can increase Ca²⁺ uptake into mitochondria via the mitochondrial Ca²⁺ uniporter. Mitochondrial Ca²⁺ overload might cause mitochondrial dysfunction, leading to reduced ATP levels and increased ROS production.⁵⁹

Given the involvement of ROS in AF pathogenesis, natural antioxidant compounds like vitamin C, vitamin E, resveratrol, or synthesized antioxidant compounds like NOX inhibitors, xanthine oxidase (XO) inhibitors, and others, have been evaluated for potential anti-AF efficacy. Many clinical studies of antioxidants have yielded mixed and inconclusive results.^60,61 Direct elimination of ROS by scavengers could be challenging, as the reaction between ROS and their targets are rapid, with the rate constant ranging from 10⁵ to 10⁹ L/mol/s, so that scavengers could miss the window to neutralize them. In addition, oxidants tend to exist in several redox states and scavengers typically can only eliminate a fraction of different forms of oxidants. Thus, targeting sources of ROS has been suggested as a better strategy.

NOXs are key enzymes that produce ROS, with NOX2 and NOX4 being the most described isoforms in cardiomyocytes. Although the NOX4-derived hydrogen peroxide production is increased in the LA of patients with AF, most studies to date are centered around NOX2.⁶² At the basal level, membrane-bound NOX2 is the main source of superoxide production in human atrial myocytes.⁶³ Both NOX2 activity and superoxide production are increased in atrial tissue of patients with AF.^63,64 Interestingly, overexpression of NOX2 in mice did not cause electrical and fibrotic remodeling, nor did it consistently promote proarrhythmic Ca²⁺ release via ryanodine receptor 2 (RyR2) cardiac calcium channels.⁶⁵ It has been suggested that NOX2-dependent ROS contributes to the local control of inositol-1,4,5-trisphosphate receptor-mediated Ca²⁺ release in atrial myocytes.⁶⁶ Apocynin, a commonly used NOX2 inhibitor, can reverse the enlarged LA, improve conduction velocity, reduce fibrosis, and suppress AF inducibility in diabetic rabbits.⁶⁷ Whether novel NOX2 blockers with improved potency and bioavailability, such as GSK2795039 and GLX481304,⁶⁸ can control ROS production and prevent AF development requires further study.

XO is another important source of ROS in cardiomyocytes. The potential of XO inhibitors, such as allopurinol and febuxostat, to mitigate AF risk has been explored. Allopurinol improves atrial electrical remodeling by suppressing Ca²⁺/calmodulin-dependent protein kinase II activity and the expression of the Na⁺/Ca²⁺ exchanger in diabetic rats and attenuates atrial structural remodeling and fibrosis in the tachypaced canine AF model.^69,70 In a 5% random survey of Medicare Claims data, allopurinol usage for >6 months was associated with reduced AF risk⁷¹; however, febuxostat was associated with increased risk of AF compared with allopurinol in older adults.⁷² Thus, more testing is required to assess the efficacy of XO inhibitors.

TARGETING THE INFLAMMASOME

ROS have been linked to heightened inflammatory responses. Despite the well-documented association between inflammation and AF in patients, traditional anti-inflammatory medications including glucocorticoids and nonsteroidal anti-inflammatory drugs have not been effective in AF. Low-dose glucocorticoids offer limited benefits, reducing post–coronary artery bypass graft surgery AF, with nontrivial risks of off-target side effects.⁷³ The role of NACHT, LRR, and PYD domain-containing protein-3 (NLRP3)-inflammasome signaling in AF pathogenesis has recently gained significant attention. Enhanced activity of the NLRP3-inflammasome was observed in atrial tissue of patients with a history of AF or those susceptible to postoperative AF.^74,75 Moreover, increased inflammasome activity is connected to obesity- or chronic kidney disease–induced atrial arrhythmogenesis.^76,77 Cardiomyocyte-specific activation of NLRP3 leads to aberrant RyR2-mediated Ca²⁺ release, shortened AERP, and increased AF inducibility in mice.⁷⁵ There is growing interest in exploring strategies to suppress NLRP3-inflammasome activity by either preventing activation or eliminating downstream effectors after activation. Currently, selective NLRP3-inflammasome inhibitors like inzomelid and dapansutrile are undergoing clinical trials to treat autoimmune/auto-inflammatory diseases. To mitigate the effect of inflammasome activation, caspase-1 inhibitors (eg, Belnacasan, VX-765) and anti-IL-1β antibodies (eg, canakinumab) have been tested in patients at high risk for heart failure with reduction in major cardiac events.⁷⁸ MCC950, a lead compound designed to block NLRP3-inflammasome activation, has demonstrated reduction in pacing-induced AF in mice.⁷⁵ Genetic ablation of NLRP3 prevents abnormal Ca²⁺ release, AERP shortening, and fibrosis—all factors associated with AF development,^75,76 positioning selective NLRP3 inhibitors as potential therapeutic agents. Because ROS can promote the polymerization of the inflammasome platform, antioxidants may also prevent inflammasome activation. Interestingly, colchicine, a well-known anti-inflammatory agent, displays inflammasome-inhibitory properties, which could be attributed to its modulation of ROS or microtubule polymerization.⁷⁹ However, although colchicine reduced postoperative AF in a few clinical studies, in the most recent international multicenter randomized COP-AF (Colchicine for the Prevention of Perioperative Atrial Fibrillation in Patients Undergoing Thoracic Surgery) trial, colchicine failed to reduce the incidence of perioperative AF.⁸⁰

Other drugs with pleiotropic effects might benefit patients with AF due to their anti-inflammatory and antioxidant properties including statins and sodium-glucose transporter 2-inhibitors. Large-scale trials and meta-analyses have provided conflicting findings on the benefits of statins for AF.⁸¹ Another approach to targeting inflammatory signaling is the promotion of active resolution of inflammation by certain mediators like resolvins, maresins, and protectins. Resolvin D1 has been shown to prevent the development of substrate for AF in rat models of right heart disease⁸² and myocardial infarction,⁸³ while suppressing active inflammation.

EARLY INTERVENTION TO PREVENT AF PROGRESSION

Underlying the “AF begets AF” paradigm is the notion that AF-induced remodeling, via a number of mechanisms, promotes ACM, which supports the perpetuation and progression of AF. Indeed, AF frequently progresses to more advanced forms during the natural history of the disease. Accordingly, it has been suggested that earlier and more aggressive management designed to maintain sinus rhythm might improve outcomes.⁸⁴ Support for this concept has emerged from a multicenter, randomized trial of early vs standard therapy of AF.⁸⁵ Patients randomized to rhythm control had a 20% lower occurrence rate of the primary outcome (composite of cardiovascular death, stroke or hospitalization for heart failure or acute coronary syndrome) than patients assigned to standard therapy. Within this study, >90% of early-intervention patients were initially treated with antiarrhythmic drugs and even after 2 years in the trial, only 19.4% of patients had received an ablation. In a more recent randomized study, early atrial cryoablation therapy was shown to be superior to early antiarrhythmic drug therapy in terms of progression to persistent AF, recurrent atrial tachyarrhythmia, AF burden, and hospitalization.⁸⁶ But another recent study⁸⁷ assessed early AF ablation vs initial pharmacological sinus-rhythm maintenance therapy followed by planned AF ablation 12 months after recruitment, finding no difference in atrial arrhythmia recurrence or AF burden between the groups, indicating that initial antiarrhythmic therapy followed by subsequent AF ablation is a perfectly viable approach.

Thus, there is a potential role for antiarrhythmic drug therapy to prevent AF and ACM development. Importantly, the use of these agents to allow for the time-dependent impact of other interventions, such as risk factor modification, to manifest could be an effective multipronged therapy. The available antiarrhythmic drug armamentarium is fairly limited, with the most recently introduced agent (dronedarone) having been introduced 14 years ago and all other agents at least 20 years ago.⁸⁸ A number of strategies have been suggested for the development of novel agents, including drugs specifically targeting atrial-selective ion channels, development of more AF-selective channel blockers, and the use of multiple ion channel-blocking agents (based on the superior effectiveness of amiodarone).⁸⁹ A major limitation to antiarrhythmic drug development has been concern about ventricular proarrhythmia. A desirable drug development approach is to produce drugs that target atrial-selective potassium channels involved in atrial repolarization. To date, this approach has been largely unsuccessful because drugs selectively targeting channels like the ultra-rapid delayed rectifier (I_Kur), the acetylcholine-dependent potassium channel (I_KACh), and calcium-dependent potassium channels (I_SK) have shown insufficient efficacy in prolonging atrial refractoriness or suppressing AF.⁹⁰ It is theoretically possible to exploit state-dependent sodium channel blockade to produce drugs that produce strong anti-AF actions on the atria at the rapid AF rates with minimal ventricular sodium channel blockade at sinus rhythm rates.⁹¹ AF selectivity can be enhanced by combining sodium channel blockade with inhibition of a potassium channel like I_Kur or the delayed rectifier.⁹² Given the perceived limited role for antiarrhythmic drugs in rhythm-control therapy, it remains to be seen whether new ion channel–based antiarrhythmic drugs for AF will be developed.

CONCLUSIONS

Contemporary ablation trials have not identified approaches that improve AF outcomes compared with PVI alone. Given the lack of progress with ablation approaches, it is essential to consider whether non-ablative adjunctive avenues are available to improve outcomes. The mechanisms underlying development of AF are multifactorial but are subject to intervention that may affect ACM (Central Illustration). Clinical and basic science data show that other approaches that specifically target the ACM underlying AF—structural, electrical, metabolic— may serve as adjunctive therapy that can lead to improved outcomes (Figure 1, Table 1). To date, various clinical trials (Table 2) provide insight into interventions that may be clinically meaningful. Future studies focused on prevention or reversal of ACM via individualized approaches could provide meaningful improvements in AF outcomes beyond what can be achieved with ablation alone.

TABLE 1.

AF Substrate-Targeting Interventions in the Treatment of AF

Therapeutic Approaches
Therapeutic Strategy	Therapeutic Agents	In Vitro	Small Animal	Large Animal	Human Patients
Lifestyle modifications
Weight loss	Diet, exercise, bariatric surgery	P	P	P	P
Moderation of alcohol	Dietary intake		P		P
Obstructive sleep apnea	Nighttime continuous positive airway pressure				I
Atrial fibrosis
ACE inhibitors	Enalapril, captopril, lisinopril	I		P	Pa
Angiotensin receptor blockers	Losartan, valsartan	P	P	P	Pa
Aldosterone antagonist	Spironolactone, eplerenone	P	P		Pa
Beta-blockers	Metoprolol	P	P	I	I
TGF-β1 inhibitors	Pirfenidone, nintedanib	I	P	P
TNF-α inhibitors	Infliximab		P		I
Antioxidants	MitoTEMPO	P	P
Oxidative stress
Antioxidants	Vitamin C, vitamin E, resveratrol	I	I	P	N
NOX2 inhibitors	GSK2795039, GLX481304	I	I
Xanthine oxidase inhibitors	Allopurinol, febuxostat	I	I	I	N
Inflammasome
NLRP3-inflammasome inhibitors	Inzomelid, dapansutrile		P
Small-molecule NLRP3 inhibitors	Compound MCC950	P	P
Tubulin disruption	Colchicine	P	I		I
Caspase-1 inhibitors	Belnacasan (VX-765)
Anti-IL-1β antibody therapy	Canakinumab				N
SGLT2 inhibitors	Empagliflozin, dapagliflozin	P	I		I
Polyunsaturated fatty acids	Resolvins, maresins, protectins		I
Early interventions
Pulmonary vein ablation	Radiofrequency or cryo ablation			P	P
Antiarrhythmic drugs	Amiodarone, sotalol, flecainide	P	P	P	P

Level of Evidence: I, inconclusive; N, negative; P, positive; blank, insufficient evidence.

^aPrimary prevention only.

ACE = angiotensin-converting enzyme; AF = atrial fibrillation; NLRP3 = NACHT, LRR, and PYD domain-containing protein-3; NOX2 = nicotinamide adenine dinucleotide phosphate oxidase enzyme 2; SGLT2 = sodium-glucose cotransporter 2; TGF = transforming growth factor; TNF = tumor necrosis factor.

TABLE 2.

Selected Clinical Trials of Interventions Targeting the AF Substrate

Study	n	Design	Clinical Setting	Treatment Group	Control Group	Follow-Up	Clinical Outcome Results
Alcohol abstinence	140	Multicenter, prospective, open-label RCT	Adults consuming ≥10 drinks (12 g alcohol), and with paroxysmal or persistent AF	Alcohol abstinence	Usual alcohol consumption	6 mo	Abstinence group had lower AF recurrence, lower AF burden, and longer time to recurrence vs the control group.
ARREST-AF	281	Nonrandomized blinded clinical trial	Post-PVI ablation, BMI ≥17 and ≥1 cardiac risk factor	RFM	Patients who declined RFM	2 y	RFM reduces AF frequency, duration, and symptoms. Single- and multiple-treatment event-free survival.
Bariatric surgery	220	Retrospective cohort study	440 consecutive morbidly obese (BMI ≥40 kg/m2) patients with AF and undergoing bariatric surgery or weight management program	Bariatric surgery	Medical weight management	1 y	Weight loss was greater in the bariatric surgery cohort, and was associated with improvement in AF type and inflammatory biomarkers.
CHARM	7,601	Multicenter RCT	Symptomatic heart failure with reduced or preserved LV function and NYHA functional class II-IV symptoms	Oral candesartan	Placebo	37.7 mo median	Treatment with candesartan reduced AF incidence in a large broadly based population of patients with symptomatic heart failure.
COP-AF	3,209	Multicenter RCT	Age >55 y undergoing noncardiac thoracic surgery	Oral colchicine	Placebo	14 d	No significant reduction in AF or myocardial injury. Increase in noninfectious diarrhea.
EARLY-AF	303	Multicenter RCT	Initial rhythm strategy for patients with paroxysmal AF	Cryoballoon PVI ablation	Class I or III antiarrhythmic drug	3 y	Cryo-PVI has lower AF recurrence and hospitalization rates. No differences noted in serious adverse events.
EAST-AFNET4	2,789	Parallel-group, open, blinded outcome assessment trial	Recent diagnosis of AF ≤12 mo before trial enrollment, and with 1+ underlying cardiovascular risk factors	PVI or antiarrhythmic drug	Usual therapy	5.1 y average	Early rhythm control had reduced adverse cardiovascular outcomes.
EMPHASIS-HF	1,794	Retrospective analysis of multicenter RCT	Heart failure patients with NYHA functional class II symptoms and LVEF ≤35%	Eplerenone	Placebo	21 mo median	New AF or atrial flutter were reduced by half in patients receiving eplerenone.
ERADICATE-AF	302	Multicenter RCT	Symptomatic paroxysmal AF	RD + PVI	PVI	1 y	RD added to PVI improves AF-free survival at 1 y.
LEAF	65	Single-center RCT	BMI ≥27 kg/m2 who opted for catheter ablation to treat AF	RFM + Liraglutide	RFM alone	6 mo	Weight loss reduces AF recurrence in obese patients with AF.
LEGACY	355	Single-center cohort study	BMI ≥27 kg/m2 and symptomatic paroxysmal or persistent AF; 355 patients agreed to physician-guided weight loss program	Three weight loss groups were compared: 1 (≥10%), 2 (3%-9%), and 3 (<3%)	None; all groups were enrolled in the program	5 y	Long-term sustained weight loss is associated with significant reduction of AF burden and maintenance of sinus rhythm.
Perioperative rosuvastatin	1,922	Multicenter RCT	Patients scheduled to undergo coronary artery bypass surgery or surgical aortic valve replacement, in sinus rhythm, no AADs	Perioperative rosuvastatin	Placebo	5 d postoperative	Rosuvastatin did not result in changes in perioperative AF vs placebo, and there were no beneficial effects on secondary outcomes.
REVERSE-AF	355	Single-center cohort study; a retrospective sub-analysis of LEGACY study	BMI ≥27 kg/m2 and symptomatic paroxysmal or persistent AF; 355 patients agreed to physician-guided weight loss program	Three weight loss groups were compared: 1 (≥10%), 2 (3%-9%), and 3 (<3%):	None; all groups were enrolled in the program	5 y	Weight loss is associated with reversal of persistent to paroxysmal AF.
SGLT2i and arrhythmias	63,166 (32 trials)	Meta-analysis	32 double-blind trials comparing SGLT2i with placebo or active control for adults with diabetes or heart failure	SGLT2i	Placebo or active control	0.46–5.7 y	SGLT2is are associated with significantly reduced risks of incident atrial arrhythmias and sudden death in diabetes.
SOLVD retrospective	374	Retrospective analysis of multicenter RCT	Heart failure with chronic systolic dysfunction, LVEF ≤35%	Oral enalapril	Placebo	2.9 y	Enalapril treatment reduced AF incidence in patients with LV dysfunction.
TRACE	1,749	Double-blind multicenter RCT	Patients with reduced LV function secondary to acute myocardial infarction	Oral trandolapril	Placebo	2–4 y	Development of AF and time to AF were significantly reduced in patients receiving trandolapril vs placebo.
VAL-HeFT	5,010	A retrospective sub-analysis of multicenter RCT	Heart failure with NYHA functional class II-IV symptoms, taking GDMT, LVEF <40%, and LVIDD/BSA >2.9 cm/m2	Oral valsartan	Placebo	23 mo mean	Occurrence of AF was lower in patients taking valsartan (5.12%) vs patients on placebo (7.95%), a 37% total reduction.

AAD = antiarrhythmic drugs (other than beta-blockers); AF = atrial fibrillation; ARREST-AF = Aggressive Risk Factor Reduction Study for Atrial Fibrillation; BMI = body mass index; CHARM = Candesartan in Heart Failure Assessment of Reduction in Mortality and Morbidity; COP-AF = Colchicine for the Prevention of Perioperative Atrial Fibrillation in Patients Undergoing Thoracic Surgery; EARLY-AF = Early Aggressive Invasive Intervention for Atrial Fibrillation; EAST-AFNET4 = Early Treatment of Atrial Fibrillation for Stroke Prevention Trial 4; EMPHASIS-HF = Eplerenone in Mild Patients Hospitalization and Survival Study in Heart Failure; ERADICATE-AF = Evaluate Renal Artery Denervation in Addition to Catheter Ablation to Eliminate Atrial Fibrillation; GDMT = guideline-directed medical therapy; LEAF = Liraglutide Effects in Atrial Fibrillation; LEGACY = Long-Term Effect of Goal-directed weight management on Atrial Fibrillation Cohort: A 5-Year follow-up study; LV = left ventricle; LVEF = left ventricular ejection fraction; LVIDD/BSA = echocardiographically measured LV index diastolic diameter/body surface area; PVI = pulmonary vein isolation; RCT = randomized controlled trial; RD = renal denervation; REVERSE-AF = PREVEntion and regReSsive Effect of weight-loss and risk factor modification on Atrial Fibrillation; RFM = risk factor management; SGLT2i = sodium-glucose cotransporter 2 inhibitors (such as: canagliflozin, dapagliflozin, empagliflozin, or ertugliflozin); SOLVD = Studies of Left Ventricular Dysfunction; TRACE = Trandolapril Cardiac Evaluation; VAL-HeFT = Valsartan Heart Failure Trial.

HIGHLIGHTS.

• As a therapeutic target for AF, atrial cardiomyopathy is not addressed by ablation.

• Biological and clinical data identify adjunctive approaches to AF management that can improve outcomes.

• Further studies personalizing therapy to individual causative factors and mechanistic mediators are needed.

ABBREVIATIONS AND ACRONYMS

ACM

atrial cardiomyopathy

ACEI

angiotensin-converting enzyme inhibitors

AERP

atrial effective refractory period

AF

atrial fibrillation

BMI

body mass index

EAT

epicardial adipose tissue

GLP-1R

glucagon-like peptide 1 receptor

IL

interleukin

LA

left atrium, left atrial

NLRP3

NACHT, LRR, and PYD domain-containing protein-3

NOX

nicotinamide adenine dinucleotide phosphate oxidase enzyme

PVI

pulmonary vein isolation

ROS

reactive oxygen species

TGF

transforming growth factor

TNF

tumor necrosis factor

XO

xanthine oxidase