Cancer-Associated Arrhythmia
Definition
Cardiac arrhythmias in patients with cancer or in cancer survivors, arising from the direct electrophysiological effects of anticancer therapy, from secondary cardiotoxicity (heart failure, ischaemia, myocarditis), from electrolyte abnormality, or from the cancer itself. Atrial fibrillation is the most common, but QT prolongation, ventricular arrhythmia and bradyarrhythmia occur across all major drug classes.
This page holds the drug-by-drug framework: incidence by agent, mechanisms, the interactions that constrain antiarrhythmic choice, QT assessment and management, ventricular arrhythmia and conduction disease. The AF management decision — rate and rhythm control, anticoagulation and the TBIP framework — is covered in concepts/AF-Special-Populations.
Key Concepts
Epidemiology
- The cancer–AF relationship is bidirectional: cancer patients have more AF and AF patients have more cancer, driven by shared risk factors of age, inflammation and metabolic disease. A cohort of more than 15,000 patients confirmed higher prevalent AF in cancer than in non-cancer controls (sources/arrhythmia-cardio-oncology-aha-2021, rating: very high).
- AF occurs in 2–16% of patients during active cancer therapy, with post-lung-surgery the highest at 6–32% (sources/Cardio-Oncology-ESC-2022, rating: very high).
- Cancer-associated AF carries 2× the stroke and thromboembolic risk and 6× the HF risk of AF in the non-cancer population (sources/Cardio-Oncology-ESC-2022).
- Incidence by agent, with the mechanism that generates it:
| Drug class | Agent | AF incidence | Mechanism |
|---|---|---|---|
| Anthracyclines | Doxorubicin | 10.3%; up to 56.6% if LVD | Secondary to cardiomyopathy |
| Alkylating agent / HSCT | Melphalan | 11% (melphalan-based) | Direct plus HSCT-related |
| BTK inhibitor | Ibrutinib | 3.5–16%; RR 4.69 (95% CI 2.17–7.64) | BTK/Tec/PI3K; calcium handling |
| BTK inhibitor (2nd generation) | Acalabrutinib | 4.1% | Higher BTK selectivity |
| BTK inhibitor (3rd generation) | Zanubrutinib | 2% | Higher BTK selectivity |
| Immune checkpoint inhibitor | Nivolumab, pembrolizumab | 25–30% in myocarditis cohorts | Myocarditis-driven |
| CAR-T | Tisagenlecleucel, axicabtagene | 7.5% | Cytokine release syndrome |
| BCR-ABL TKI (3rd generation) | Ponatinib | 1.9–7% | Unclear |
| VEGF inhibitor | Sorafenib + 5-FU | 5.1% (small series) | Unclear |
(sources/arrhythmia-cardio-oncology-aha-2021)
Mechanisms — direct versus secondary arrhythmogenesis
- Ibrutinib carries the highest AF burden of any targeted agent: relative risk 4.69 (95% CI 2.17–7.64, P<0.001) in a meta-analysis of 8 RCTs (n=2,580) (sources/arrhythmia-cardio-oncology-aha-2021, rating: very high).
- Proposed mechanisms are both on- and off-target: BTK inhibition, off-target Tec kinase inhibition, PI3K pathway inhibition, and enhanced automaticity from altered sarcoplasmic reticulum calcium handling.
- Patients who develop AF on ibrutinib have higher mortality, probably through ischaemic or haemorrhagic cerebrovascular events rather than arrhythmic death.
- Newer-generation agents have lower AF rates through greater BTK selectivity, and pirtobrutinib (third-generation, reversible) reports 3.2%; the choice between agents should weigh individual AF risk. (sources/arrhythmia-cardio-oncology-aha-2021, sources/cancer-rx-cv-acc-2025)
- Most cancer-drug arrhythmia is secondary, not direct. Anthracycline-associated AF follows cardiomyopathy and atrial remodelling; AF in checkpoint-inhibitor patients occurs mainly within myocarditis cohorts; CAR-T arrhythmia follows cytokine release. The exceptions are QT prolongation, which is a direct channel effect, and ibrutinib's ventricular arrhythmia (see below).
Drug interactions that constrain antiarrhythmic choice
Rate and rhythm control in these patients is limited less by efficacy than by metabolism. CYP3A4 and P-glycoprotein dominate, and combinations must be checked drug by drug rather than assumed safe by class (sources/cardio-oncology-drug-aha-2022, rating: very high).
| Cardiovascular drug | Interaction with ibrutinib | Action |
|---|---|---|
| Diltiazem, verapamil | Raise ibrutinib 6–9× through CYP3A4 inhibition | Avoid for rate control; use a beta-blocker |
| Amiodarone, dronedarone | Raise ibrutinib 6–9× through CYP3A4 inhibition | Avoid for rhythm control; if unavoidable, reduce or withhold ibrutinib |
| Dabigatran | Raised by P-glycoprotein inhibition | Contraindicated; use a factor Xa inhibitor |
| Digoxin | Raised by P-glycoprotein inhibition | Avoid |
| Amlodipine, nifedipine | Dihydropyridine CCBs are CYP3A4 substrates but do not inhibit ibrutinib metabolism | Safe |
(sources/cancer-rx-cv-acc-2025, rating: very high)
Dronedarone deserves separate caution in cancer beyond the ibrutinib interaction: its combined CYP3A4 and P-glycoprotein effects cause dangerous accumulation with several anticancer agents.
Atrial fibrillation — rate and rhythm control
The full decision framework, including the TBIP anticoagulation algorithm and agent selection by cancer type, is in concepts/AF-Special-Populations. What is specific to this page:
- Rate control with a beta-blocker is preferred. Non-dihydropyridine calcium channel blockers are contraindicated with ibrutinib, and digoxin should be avoided in the same patients (see the table above).
- The ibrutinib-AF algorithm starts with haemodynamic stability: unstable patients are cardioverted; stable patients are assessed for reversible causes and rate-controlled with a beta-blocker as first choice, anticoagulated with a factor Xa inhibitor in preference to dabigatran, with left atrial appendage closure and AF ablation acceptable in selected patients without an oncological alternative (sources/cancer-rx-cv-acc-2025, rating: very high).
- Rhythm control is less well supported. The benefit of early rhythm control in EAST-AFNET 4 has not been tested in cancer patients, who were not enrolled; cardioversion is less durable in an inflammatory cancer milieu; and antiarrhythmic drugs interact with anticancer therapy and require pharmacological review before use (sources/arrhythmia-cardio-oncology-aha-2021, rating: very high).
Anticoagulation
- CHA₂DS₂-VASc remains the recommended tool but probably underestimates thromboembolic risk in cancer, because cancer-specific predictors — type, stage, active treatment — are not represented (sources/arrhythmia-cardio-oncology-aha-2021, rating: very high).
- HAS-BLED underestimates bleeding risk because it captures neither thrombocytopenia nor intracranial metastases. Cancer-specific bleeding algorithms are needed and do not exist.
- DOACs are preferred over warfarin for non-valvular AF: the ARISTOTLE cancer subgroup (n=1,236) confirmed apixaban superior to warfarin for stroke and systemic embolism, and ENGAGE AF-TIMI 48 showed similar findings for edoxaban.
- Aspirin is not a substitute for anticoagulation (Class III); thrombocytopenia and ibrutinib-related platelet dysfunction make bleeding more likely without offering reliable protection.
- TBIP — thrombotic risk, bleeding risk, drug–drug Interactions, Patient preference — is the ESC framework that organises the decision, with thresholds, the very-high-bleeding exclusions and the LMWH exceptions set out in concepts/AF-Special-Populations (sources/Cardio-Oncology-ESC-2022, rating: very high).
QT prolongation
- Incidence reaches 22% across cancer drug classes; arsenic trioxide is highest at 26–93% in acute promyelocytic leukaemia. Life-threatening arrhythmia remains rare (<1%) despite these rates, so interrupting cancer therapy for QT prolongation is a joint oncology, cardiology and patient decision rather than an automatic one (sources/arrhythmia-cardio-oncology-aha-2021, rating: very high).
- Mechanisms: direct potassium channel blockade, and PI3K pathway inhibition by many TKIs, both delaying ventricular repolarisation.
- Measurement: manual calculation in lead II or V5, averaged over 3–5 beats where there is sinus arrhythmia or 10 beats in AF. Fridericia (QTcF) is preferred to Bazett in this population because it is more accurate at heart rate extremes. The JT interval can substitute with bundle branch block or ventricular pacing but is difficult to apply clinically.
- Monitoring: no standardised protocol exists — follow drug-label instructions and obtain an ECG after the first dose and after each dose change for novel agents.
- Arrhythmic risk threshold: QTc >500 ms, or a change from baseline >60 ms.
- Acute management: stop non-essential QT-prolonging drugs, correct potassium and magnesium, give IV magnesium sulfate for torsades, maintain heart rate >100 bpm with isoproterenol or temporary pacing, and use ACLS with lidocaine for refractory cases. See concepts/Torsades-de-Pointes for general management.
- Mexiletine has emerging evidence for arsenic-induced QT prolongation and recurrent torsades; RCT data for broader cancer use are lacking.
Ventricular arrhythmia
- Sustained VT or VF from chemotherapy is relatively rare and more frequent in advanced metastatic disease (sources/arrhythmia-cardio-oncology-aha-2021, rating: very high).
- Ibrutinib-associated ventricular arrhythmia occurs independently of QT prolongation — ibrutinib may in fact shorten the QT interval — and is attributed to enhanced automaticity and triggered activity. The mechanism is not characterised well enough to identify which patients are at risk.
- Anthracycline-associated ventricular arrhythmia is secondary to LV systolic dysfunction and electrolyte abnormality, particularly hypokalaemia, rather than a direct proarrhythmic drug effect.
- ICD indications follow standard primary prevention criteria — LVEF ≤35%, NYHA II–III, life expectancy >1 year. Chemotherapy-induced cardiomyopathy is not excluded from ICD indication.
- MADIT-CHIC (single-arm, prospective, multicentre) studied CRT in chemotherapy-induced cardiomyopathy with LVEF ≤35% and LBBB (median QRS 152 ms) on optimal medical therapy: LVEF improved, LV and LA volumes fell, and symptoms improved at 6 months. CRT-P, not only CRT-D, is appropriate even when life expectancy is under 1 year, for symptom palliation rather than survival.
Bradyarrhythmia and conduction disease
| Drug | Bradycardia incidence | Notes |
|---|---|---|
| Paclitaxel | Up to 30% (asymptomatic sinus) | Reversible; advanced forms atypical |
| Thalidomide | Up to 40% sinus | Some require pacemaker |
| 5-Fluorouracil | ~12% symptomatic (one series) | Lower in systematic reviews |
| Ibrutinib | 2.3% conduction defects | 42% high-grade or complete AV block; 18% fatal |
| Crizotinib, ceritinib | Common but mild | Rare symptomatic episodes |
| ICI (nivolumab and others) | 17% conduction disorders in a cardiotoxicity cohort | May be the first manifestation of myocarditis |
(sources/arrhythmia-cardio-oncology-aha-2021, rating: very high)
- ICI-associated AV block may be the first sign of myocarditis (about 1% overall, 50% mortality); in one series conduction disorders carried 80% versus 16% cardiovascular mortality. Any new conduction disease in a checkpoint-inhibitor patient should prompt assessment for myocarditis rather than isolated pacing.
- Screen for sleep-disordered breathing before implanting (Class I): nocturnal bradycardia from untreated apnoea reverses with CPAP, so the pacing decision should follow assessment rather than precede it (sources/sdb-arrhythmia-aha-2022, rating: very high).
- Reduce AV-nodal blocking drugs pre-emptively before starting a bradycardic anticancer agent — taxanes, ALK inhibitors or ibrutinib.
Contradictions / Open Questions
- True incidence is unknown. Dedicated prospective studies of arrhythmia in cancer are lacking, most evidence is retrospective, and the reported rates probably underestimate the burden (sources/arrhythmia-cardio-oncology-aha-2021, rating: very high).
- BTK inhibitor AF rates differ between statements. The AHA 2021 statement gives acalabrutinib 4.1%, zanubrutinib 2% and ibrutinib 3.5–16%, while the 2025 ACC guidance gives 5–9%, 3–6% and 3–16% respectively; the two draw on different trial sets and registries, and neither is validated in an unselected cancer population (sources/cancer-rx-cv-acc-2025).
- Early rhythm control is untested in cancer. EAST-AFNET 4 excluded these patients, so whether its benefit transfers, and how it interacts with a limited cancer prognosis, is unknown.
- The mechanism of ibrutinib ventricular arrhythmia is uncharacterised — enhanced automaticity versus triggered activity has not been established, and no risk-stratification tool follows from it.
- Prophylactic antiarrhythmic therapy before HSCT has never been tested in an RCT, despite correlational data linking pre-existing arrhythmia to post-transplant mortality.
- Standard stroke and bleeding scores are not validated in cancer, and the cancer-specific algorithms repeatedly called for do not yet exist.
Connections
The AF decision itself
- concepts/AF-Special-Populations — rate and rhythm control, TBIP, agent choice and bleeding thresholds in cancer-AF
Arrhythmia mechanisms and management
- concepts/Torsades-de-Pointes — general management of drug-induced torsades
- concepts/Drug-Induced-Arrhythmia — drug-induced arrhythmia outside oncology
- entities/Atrial-Fibrillation — the arrhythmia itself
Cardio-oncology domain
- concepts/Cardio-Oncology — the discipline and care pathways
- concepts/Cancer-Therapy-Related-CV-Toxicity — drug profiles, monitoring and ICI myocarditis management
- concepts/Autonomic-Dysfunction-in-Cancer — the other neuro-cardiac complication of cancer therapy
Rhythm devices
- entities/ICD — primary prevention indications applied here unchanged
- concepts/Cardiac-Resynchronization-Therapy — the intervention tested in MADIT-CHIC
- entities/Heart-Failure — the substrate for most secondary arrhythmia
Sources
Arrhythmia in cancer
Guidelines and drug interactions
Bradycardia and pacing