Atrial fibrillation (AF) is common in patients with malignancy due to shared risk factors (age, inflammation, metabolic disorders) and cardiotoxic cancer therapies. In this population, long-term oral anticoagulation (OAC) is difficult to sustain because hypercoagulability coexists with bleeding tendencies from thrombocytopenia, mucosal disease, and treatment-related organ dysfunction. Balancing stroke prevention against bleeding risk is therefore uniquely challenging.
Catheter ablation (CA) is an established rhythm-control option for symptomatic, drug-refractory AF, while left atrial appendage occlusion (LAAO) provides a mechanical alternative to OAC for stroke prevention. Performing one-stop CA + LAAO in a single session may minimize hospitalizations and anesthesia exposure, shorten the periprocedural anticoagulation window, and reduce disruptions to oncologic care. Yet evidence specific to oncology is scarce. Findings from general AF cohorts—suggesting procedural efficiency and non-inferior safety—may not extrapolate because cancer-related hematologic instability, systemic inflammation, and drug interactions can alter risks of tamponade, major bleeding, device-related thrombus (DRT), and arrhythmia recurrence.
To address this gap, we compared one-stop CA + LAAO with a staged approach in cancer patients with non-valvular AF using propensity-score matching. Outcomes included 30-day peri-procedural safety (composite of all-cause death, tamponade, BARC ≥3 bleeding, DRT) and one-year effectiveness (freedom from AF recurrence off antiarrhythmics and absence of ischemic stroke/TIA). We hypothesized that the one-stop strategy would reduce fluoroscopy time and length of stay without compromising safety or mid-term protection.

Hospital records from January 2023 to June 2025 were reviewed. Patients who underwent CA and LAAO during the same admission formed the one-stop group (n = 35); those who received CA followed by LAAO within 12 months comprised the staged group (n = 35). Exclusion criteria were left-atrial-appendage thrombus, expected survival < 12 months, or follow-up < 12 months. One-to-one nearest-neighbour propensity-score matching balanced age, sex, CHA₂DS₂-VASc score, HAS-BLED score, cancer type and stage (standardised mean difference < 0.10 for all covariates).The primary safety endpoint was a composite of all-cause death, cardiac tamponade, BARC ≥ 3 bleeding, or device-related thrombosis (DRT) within 30 days.The primary effectiveness endpoint was freedom from documented AF recurrence and ischaemic stroke/TIA at 12 months off anti-arrhythmic drugs. Categorical variables were analysed with Fisher’s exact test; absolute risk differences (ARD) with 95 % confidence intervals (CI) are reported.

】After matching, mean CHA₂DS₂-VASc and HAS-BLED scores were 4.1 ± 1.2 and 3.5 ± 0.8, respectively; 61 % of patients had solid tumours (mainly lung 18 %, breast 15 %, colorectal 10 %) and 39 % had haematological malignancies.Procedural metrics.Fluoroscopy time: 10 ± 3 min (one-stop) vs 17 ± 4 min (staged), P < 0.001.Length of stay: median 2 (IQR 2–3) days vs 4 (3–5) days, P < 0.001.Primary safety endpoint (30 days).One-stop 1/35 (2.9 %) vs staged 3/35 (8.6 %); ARD = -5.7 % (95 % CI –16.2 % to 4.8 %); Fisher P = 0.62.Major bleeding:1 vs 2 patients.DRT: 0 vs 1 patient.Primary effectiveness endpoint (12 months).One-stop 26/35 (74.3 %) vs staged 25/35 (71.4 %); ARD = 2.9 % (95 % CI –14.0 % to 19.8 %), P = 0.78. Ischaemic stroke/TIA: 1.8 % vs 1.8 % (each 1 patient).No significant interaction was observed between treatment strategy and cancer type (solid vs haematologic; P\_interaction = 0.67).

In this single-centre exploratory analysis, the one-stop CA-LAAO strategy significantly reduced fluoroscopy time and hospital stay without increasing major peri-procedural complications. One-year rhythm control and stroke prevention were comparable with the staged approach. Although limited by small sample size, these findings suggest that a one-stop procedure can lessen treatment burden in cancer patients with AF and merits confirmation in larger prospective studies.