Biological Therapies for Atrial Fibrillation
Definition
Biological therapies for AF are molecular interventions — gene transfer, RNA manipulation (miRNA overexpression/suppression), and cell-based therapy — that target the underlying atrial substrate perpetuating AF, rather than transiently suppressing ion channels (drugs) or irreversibly destroying tissue (ablation). The goal is durable modification of the electrical, structural, or autonomic substrate driving AF progression.
Key Concepts
Overview: Preclinical Evidence
As of 2018, 25 placebo-controlled preclinical animal studies have been conducted (McRae et al., Heart Rhythm 2019; systematic review). All 25 interventions improved at least one primary outcome vs controls:
- AF inducibility: 85% reduction (OR 0.15; 95% CI 0.07–0.35; P<0.01)
- Atrial fibrosis: −6.7% absolute content (95% CI 4.2–9.2; P<0.01)
- Days in sinus rhythm: +6.4 days (95% CI 5.83–6.97; P<0.01)
No biological therapy has yet entered clinical trials. (sources/biological-tx-af-hrs-2019, rating: medium; sources/gene-therapy-arrhythmia-2025, rating: high)
Strategy Class 1 — Fibrosis Reduction (Most Replicated)
Atrial fibrosis is driven primarily by TGF-β signalling → myofibroblast activation → collagen deposition → conduction heterogeneity → AF maintenance.
- Anti-miR-21: Most replicated strategy (4 independent labs — Cardin, Adam, He, Huang). miR-21 upregulation in AF promotes fibroblast survival via PTEN/Sprouty suppression. Anti-miR-21 delivery (plasmid/lentivirus) reduced atrial fibrosis by ~5–10% absolute in rat/mouse/rabbit models. (sources/biological-tx-af-hrs-2019, rating: medium)
- ACE2 overexpression: Adenoviral epicardial delivery in dogs (3 studies — Fan, Zhou ×2); reduces angiotensin II → TGF-β signalling → fibrosis −6–9%; also improved conduction and reduced AF inducibility. (sources/biological-tx-af-hrs-2019, rating: medium)
- Dominant-negative TGF-β1 receptor: Plasmid, dogs with HF-induced AF substrate (Kunamalla 2016); fibrosis −8.7%; preserved conduction velocity. (sources/biological-tx-af-hrs-2019, rating: medium)
- miR-30c overexpression (AAV): Targets TGFβRII; reduces TGF-β1-induced fibroblast activation in rats. (sources/biological-tx-af-hrs-2019, rating: medium)
- miR-27b overexpression (AAV, pericardial): Targets ALK5 (TGF-β receptor I); fibrosis −15.7% in angiotensin II-infused mice. (sources/biological-tx-af-hrs-2019, rating: medium)
Strategy Class 2 — Repolarisation Modification
AF promotes electrical remodelling: ↑IKACh, ↓ICaL, ↑IK1 → shortened atrial APD/ERP → increased reentry vulnerability.
- Dominant-negative KCNH2 mutant (G627S/G628S; ↓IKr): Adenoviral delivery, pigs (Amit 2010; Soucek 2012; Liu 2017) — prolongs atrial APD and ERP; validated by 2 independent labs; effect lost at 21 days with adenoviral washout. (sources/biological-tx-af-hrs-2019, sources/gene-therapy-arrhythmia-2025, rating: medium/high)
- Anti-miR-328 (↑ICaL normalisation): Suppresses miR-328-mediated downregulation of CACNA1C; two independent labs (Lu/mice, Li/rabbits). (sources/biological-tx-af-hrs-2019, rating: medium)
- miR-26a overexpression (↓IK1): Adenovirus IV injection in mice — adverse event: QT prolongation (only adverse event in this systematic review; likely reflects off-target ventricular IK1 suppression via systemic IV delivery). (sources/biological-tx-af-hrs-2019, rating: medium)
- Anti-miR-1 (↓delayed rectifier): Lentivirus, rabbits — reduces proarrhythmic shortening of atrial ERP. (sources/biological-tx-af-hrs-2019, rating: medium)
- KCNK2 overexpression (↑TREK-1, K2P channel): Adenovirus, pigs; reduces AF inducibility and fibrosis simultaneously. (sources/biological-tx-af-hrs-2019, rating: medium)
- SERCA2a overexpression (AAV9): Rabbit burst pacing model — restores SR Ca²⁺ cycling; reduces triggered activity from Ca²⁺ overload/spark-driven NCX currents. (sources/biological-tx-af-hrs-2019, rating: medium)
Strategy Class 3 — Conduction Velocity Preservation
AF causes ↓Cx40/Cx43 expression → slowed atrial conduction → reduced wavelength → multiple simultaneous reentry wavefronts.
- GJA1 (Cx43) overexpression: Adenovirus epicardial, pigs (Igarashi 2012; Bikou 2011) — preserved conduction velocity (27→41 cm/s), reduced AF inducibility >80%, +6.6 days sinus rhythm; validated by 2 independent labs. (sources/biological-tx-af-hrs-2019, sources/gene-therapy-arrhythmia-2025, rating: medium/high)
- GJA5 (Cx40) overexpression: Adenovirus, pigs (Igarashi 2012) — similar directional effect; less robust than Cx43. (sources/biological-tx-af-hrs-2019, rating: medium)
Strategy Class 4 — Autonomic Modulation
Vagal-tone-mediated IKACh activation shortens atrial ERP and promotes AF; SNS hyperactivity in HF drives structural remodelling.
- Gαi/o inhibition (Gαi C-terminal peptide plasmid): Disrupts muscarinic receptor signalling → reduces constitutive IKACh activation in AF; dogs (Aistrup 2011). Only study with dose-response relationship. (sources/biological-tx-af-hrs-2019, sources/gene-therapy-arrhythmia-2025, rating: medium/high)
- miR-206 suppression (lentivirus): Modulates intrinsic cardiac autonomic nerve remodelling via SOD1 regulation; dogs (Zhang 2015). (sources/biological-tx-af-hrs-2019, rating: medium)
Strategy Class 5 — Anti-Apoptosis
AF-induced atrial myocyte death contributes to irreversible replacement fibrosis and conduction failure.
- siRNA caspase-3 knockdown: Plasmid, pigs (Trappe 2013); reduced cardiomyocyte apoptosis, delayed progression to sustained AF, +4.3 days sinus rhythm. (sources/biological-tx-af-hrs-2019, sources/gene-therapy-arrhythmia-2025, rating: medium/high)
Cell Therapy
- Mesenchymal stem cells (IV infusion): Single study (Ramos, rats with experimental sleep apnea); reduced atrial fibrosis −2.4% via paracrine effects; the only cell therapy study in this systematic review. (sources/biological-tx-af-hrs-2019, rating: medium)
Delivery Systems and Translational Barriers
| System | Durability | Delivery route | Status |
|---|---|---|---|
| Naked plasmid | 3–4 days | IM/epicardial direct | Proof-of-concept only |
| Adenovirus | 3–21 days | IM/epicardial direct | Pre-clinical only; immunogenic |
| Lentivirus | Months–years (integrating) | IM/IV | Safety concerns (insertional mutagenesis) |
| AAV (AAV9, AAV variants) | Years (non-integrating) | IV/intraperitoneal | Most clinically translatable; limited atrial tropism |
| siRNA/ASO | Days–weeks | IV (with targeting) | Liver targeting established; cardiac targeting emerging |
- 80% of studies used direct intramyocardial or epicardial delivery — incompatible with clinical scale
- All but 2 studies had follow-up ≤21 days — biological durability untested beyond 3 weeks
- Systemic targeting: AAV with tissue-specific promoters or atrial-targeting vectors (analogous to asialoglycoprotein-targeted ASO for liver/PCSK9) needed for clinical translation
- GMP manufacturing cost explodes from mouse dose to clinical dose — non-trivial commercial barrier
Critical Translational Gaps
- No therapy delivered after AF is established: All 25 studies deliver before AF onset; there is no realistic clinical scenario for prophylactic treatment before the first AF episode
- Single-mechanism limitation: AF in HF involves ↑fibrosis + ↑SNS + ↑IKACh + ↓Cx43 simultaneously — single-target approaches likely provide only partial efficacy (cf. 81% recurrence with any single antiarrhythmic drug after cardioversion)
- Spontaneous AF models underused: Burst-pacing-induced AF models are artificial; spontaneous AF models (recently developed) better replicate the human initiator-substrate-trigger triad
- No clinically relevant endpoints: AF inducibility at EPS is a surrogate; spontaneous AF frequency/duration, atrial thrombogenesis, and heart rate control are the clinically meaningful outcomes — used in only a minority of studies
- Publication bias is nearly complete: 0/25 studies reported a failed intervention; file-drawer negative studies almost certainly exist
Emerging Targets (Not Yet Explored, 2018)
- Atrial inflammation (lymphomononuclear infiltration documented in lone AF biopsies)
- Circadian clock genes regulating ion channel transcription
- Mechanoelectric feedback (stretch-activated channels)
- Hormonal modulation of channel activity (AT1R, adrenergic crosstalk)
- Multi-target regulatory network approaches specific to HF+AF vs. lone AF vs. post-operative AF
Contradictions / Open Questions
- Universal preclinical success vs. zero clinical translation: 96% of biological interventions showed benefit in animals; no clinical trial has been initiated. This discrepancy reflects delivery barriers, pre-AF prophylaxis design, publication bias, and GMP cost — not biological implausibility. (sources/biological-tx-af-hrs-2019, rating: medium)
- Short follow-up (≤21 days) and adenoviral washout: Many studies show efficacy within 2–3 weeks, but adenoviral expression decays by 21 days. Whether effects on fibrosis and conduction persist after transgene loss is largely untested; anti-fibrotic effects may be more durable than electrophysiological effects. (sources/biological-tx-af-hrs-2019, rating: medium)
- miR-26a IV delivery → QT prolongation: The only adverse event reported in 25 studies was QT prolongation from systemic IV adenoviral delivery of miR-26a — likely reflecting off-target ventricular IK1 suppression (IK1 maintains the resting membrane potential in ventricular myocytes; suppression prolongs APD and QTc). This signals that systemic delivery of IK1-targeting agents carries pro-arrhythmic ventricular risk — atrial-selective delivery is mandatory. (sources/biological-tx-af-hrs-2019, rating: medium)
- SERCA2a in AF vs. HF: SERCA2a overexpression (CUPID trial in HF) failed the phase 2b/3 human trials despite promising preclinical HF data — the translational failure pattern extends to AF; similar disconnect should be anticipated. (sources/biological-tx-af-hrs-2019, rating: medium)
- Cell therapy underrepresented (1/25 studies): In contrast to HF, where cell therapy has been extensively studied, AF cell therapy has essentially one published report; paracrine fibrosis-reduction by MSCs is conceptually plausible but nearly untested. (sources/biological-tx-af-hrs-2019, rating: medium)
Connections
- Related to entities/Atrial-Fibrillation — primary disease target
- Related to concepts/AAV-Gene-Delivery — durable delivery platform; 4.7 kb packaging constraint relevant to multi-target strategies
- Related to concepts/Gene-Silencing-Therapy — miRNA-based suppression/overexpression strategies
- Related to concepts/Electrical-Remodeling — ionic targets of AF biological therapies (ICaL, IKr, IK1, IKACh)
- Related to concepts/Biological-Pacemaker — gene therapy approaches to cardiac ion channel modulation (adjacent concept)
- Related to entities/KCNH2 — dominant-negative KCNH2-G628S used in Amit/Soucek AF trials
Sources
- sources/biological-tx-af-hrs-2019 — systematic review + meta-analysis; 25 preclinical studies; primary source
- sources/gene-therapy-arrhythmia-2025 — broader gene therapy arrhythmia review; overlapping evidence base