This article is mentioned in:

Introduction

Atrial fibrillation (AF), the most common sustained arrhythmia, exhibits a high prevalence associated with ageing, affecting an estimated 33.5 million individuals globally (1). Typically, AF is associated with a five-fold increase in strokes, which is more devastating than stroke caused by cerebrovascular diseases or hypertension. Although most patients with AF are asymptomatic, stroke often emerges as the primary symptom of AF. Notably, ≥95% of these thrombi originate from the left atrial appendage. AF predisposes individuals to atrial and auricular thrombosis through intricate interactions among local, systemic and hemodynamic factors, significantly elevating the likelihood of cerebral and systemic thromboembolic events, thereby impacting morbidity and mortality rates. Extensive research over decades has elucidated the formation of the left atrial appendage thrombosis, primarily linked to structural abnormities, coagulation mechanisms and endothelial damage. Consequently, various therapeutic strategies have been developed to mitigate these risks (2). For patients with symptomatic AF, radiofrequency ablation is frequently employed to alleviate symptoms and prevent recurrence. However, achieving disease control without resorting to radiofrequency ablation is often the goal. Nevertheless, patients with a history of left atrial appendage thrombosis are presumed to have a heightened risk of stroke compared with others. There is increasing recognition that sole administration of oral anticoagulants after successful thrombolysis may increase the risk of stroke and death (2).

Traditionally, systemic anticoagulation has been an effective strategy for preventing atrial and left atrial thrombosis. Initially, warfarin and other vitamin K antagonists were the primary oral anticoagulants used in China until 2009. Despite their efficacy in preventing thromboembolism, their narrow therapeutic index necessitated frequent monitoring and dosage adjustments, posing a substantial risk of bleeding or thrombosis and often leading to poor patient compliance (3). Therefore, new oral anticoagulants (NOACs) were developed, demonstrating comparable efficacy and safety to warfarin in preventing stroke and systemic embolism in patients with AF. Moreover, these NOACs should not exhibit the limitations of vitamin K antagonists, especially in terms of safety. Several NOACs, including dabigatran, rivaroxaban, apixaban and edoxaban, have been developed and approved by regulatory agencies for the prevention of stroke in patients with non-valvular AF. However, their efficacy in thrombus regression remains unclear, with limited data available on their treatment of left auricular thrombosis (4). Additionally, the European Society of Cardiology (ESC) 2020 guidelines recommended anticoagulation for at least 2 months post-radiofrequency ablation based on the patient's stroke risk profile, recognising AF itself as a thrombosis factor. Therefore, reducing the AF burden may decrease the risk of ischaemic stroke (5). Nevertheless, controversy persists regarding the necessity of combined radiofrequency ablation treatment and the choice of anticoagulant therapy in patients with AF at risk of thrombotic events.

Motivated by these considerations and given the high propensity for repeat thrombotic events in patients with a history of left atrial appendage thrombosis, the present study explored the efficacy and adverse events of radiofrequency ablation combined with anticoagulation treatment in patients with AF and previous left auricular thrombosis. Moreover, the efficacy of traditional anticoagulant warfarin was compared with NOACs while analysing factors influencing patient mortality and stroke risk. Consequently, the present study aimed to enhance treatment planning for patients with AF with previous left atrial appendage thrombosis and elucidate factors contributing to their poor prognosis. By focusing on this special population, the present study endeavoured to explore the relationship between heart rhythm disorders and recurrent thrombosis, aiming to inform clinical practice rigorously.

Materials and methods

General information

A total of 92 patients (age, 66.90±8.3 years; 57 males and 35 females) diagnosed with AF and previous left atrial thrombus, who were treated at Shanghai Jiao Tong University Affiliated Chest Hospital (Shanghai, China) between March 2018 and March 2023 were retrospectively enrolled. Successful thrombolysis was confirmed via trans-esophageal echocardiography (TEE) or intracardiac ultrasound (ICE). Based on treatment criteria, patients were divided into two groups: The radiofrequency ablation (RA) combined anticoagulation group (Group A) and the simple anticoagulation and antiarrhythmic drug group (Group B). In Group A, patients underwent RA in conjugation with anticoagulant medication upon confirmation of thrombus dissolution, while Group B received oral anticoagulant and antiarrhythmic medications post-procedure. Oral medication was discontinued after three months in Group A, whereas patients in Group B continued long-term oral medication. Subgroup analyses were conducted within Groups A and B, based on left atrial internal diameter (≤45 mm), duration of AF (≤1 year) and type of anticoagulant. Patients with concomitant valvular disease were excluded. The present study was approved (approval no. IS22031) by the Ethics Committee of Shanghai Jiao Tong University Affiliated Chest Hospital (Shanghai, China) and all patients provided signed informed consent after family approval. All procedures involving human participants conformed to the ethical guidelines of the Declaration of Helsinki (2013.10 revision).

Inclusion and exclusion criteria. Inclusion criteria

Inclusion criteria were as follows: i) Patients aged ≥55 years; ii) patients diagnosed with non-valvular AF combined with left atrial thrombus, confirmed by TEE or other means and successful thrombolysis assessed by TEE or ICE; iii) patients meeting criteria for RA; and iv) patients having no allergies to the anticoagulants used in the study and voluntarily participation with signed informed consent.

Exclusion criteria. Inclusion criteria were as follows: i) patients with valvular AF, absence of left auricular thrombus and non-resolution of thrombus post-anticoagulation therapy; ii) patients with mechanical valve implantation or moderate-to-severe mitral stenosis; iii) patients with uncontrolled bleeding disorders; and iv) patients who were prescribed with anticoagulant drugs other than warfarin and the NOAC selected in the present study and/or patients with cerebrovascular or other cardiac-related diseases.

Treatment and examination. Ablation protocol

Catheter ablation was performed under the guidance of an electro-anatomical mapping system (CARTO, Biosense Webster, Inc.; Johnson & Johnson MedTech) and followed a sequential approach: i) Circumferential pulmonary vein isolation (PVI): The radiofrequency energy (RF) was 40-45 W (15-20 ml/min saline perfusion) with ablation index of 400-500 during PVI. Successful PVI was defined as eliminating all PV potentials and/or atrium-PV potential dissociation; ii) Linear ablation: The anterior roof, lateral mitral isthmus was ablated with linear lesion (RF: 40-45 W, 15-20 ml/min saline perfusion). If bidirectional block could not be achieved in the mitral isthmus, epicardial ablation within the coronary sinus or vein of Marshall ethanol infusion was performed at the operator's discretion with an error margin of ±2% (or Cronbach's alpha coefficient of 0.85) to indicate the variability in the procedure. In the RA group, linear ablation at tricuspid isthmus was performed; iii) Driver ablation: The targeted ablation regions had electrograms displayed: i) Spatial-temporal dispersion potentials, which spread over atrial fibrillation cycle length (AFCL) at a minimum of 3 adjacent bi-pole and associated with local AFCL ≤ mean AFCL potentials; ii) locally short cycle length activity; iii) focal activity. Localized ‘patch’ ablation was applied (40-45 W) at any atrial location except appendage and vena cava. If two ablated regions were very close (<1 cm), linear ablation was performed to connect the two regions; and iv) post-AF treatment: If AF converted to atrial tachycardia during the procedure, the arrhythmia was mapped and ablated until conversion to sinus rhythm (SR). After SR conversion, AF re-induction was tested by burst atrial pacing (Pacing CL: 200-250 msec). If the AF could not be terminated by ablation, 150-200 J direct current conversion was performed.

During the RA treatment, 10,000 U of normal heparin (UH) was administered before the atrial septal puncture. Furthermore, the activated clotting time (ACT) was checked 15 min after the administration and every 20 min subsequently. Moreover, the weight-adjusted UH pushes were performed to maintain ACT for 300-350 sec while the catheter remained in the left atrium.

Medication. In Group A, post-thrombus dissolution, patients underwent RA and received anticoagulant and antiarrhythmic medications, including Amiodarone and β-blocker. Conversely, Group B patients received oral anticoagulant and antiarrhythmic medications post-procedure. Oral medication was discontinued after three months in Group A, whereas patients in Group B continued long-term oral medication.

Follow-up and endpoints. Follow-up was conducted through outpatient clinical visits every 2-4 weeks for the first 3 months and every 1-3 months thereafter. Monthly telephone follow-ups were also performed until the study's conclusion. The primary endpoint was defined as cerebral embolism due to cardiogenic thrombus shedding or death. Secondary endpoints included other thrombotic events such as cerebral infarction, lower extremity venous thrombosis, mesenteric embolism or transient ischemic attacks (TIAs). The final follow-up was conducted in February 2023, with a median follow-up duration of 36.02 months.

Observational indicators

Various observational indicators were recorded, including age, sex, hypertension, medical history (diabetes mellitus), diagnostic characteristics, duration of AF, left atrial internal diameter, left ventricular ejection fraction (LVEF), dilated left ventricular end-diastolic diameter (LVEDD), CHADS, B-type natriuretic peptide (BNP), proBNP, auricular opening, auricular long axis, distal auricular, distal to the opening, type of anticoagulant and time intervals for repeat ultrasound in Groups A and B and their respective subgroups (months). Survival outcomes (stroke and death) were documented during follow-up after concurrent treatment in Groups A and B and their subgroups. Stroke diagnosis followed clinical guidelines, including symptoms and cranial imaging evidence. Univariate and multivariate logistic analyses were conducted to identify risk factors influencing stroke occurrence and death in patients with AF and left atrial thrombus.

Statistical analysis

Due to the exploratory nature of the present study, formal sample size calculation was not performed. A sample size exceeding 20 patients per group met the minimum statistical requirements. The statistical analyses were performed using the SPSS version 25.0 software (IBM Corp.). Normally distributed continuous data are expressed as the mean + standard deviation (SD). The data between the two groups were compared using the independent samples t-test. Categorical variables were compared using χ2 tests or Fisher's exact test, as appropriate. Skewed distribution data were expressed as a median and interquartile range, with comparisons between groups performed using the Wilcoxon rank sum test. Additionally, categorical data were expressed as cases (%) and analysed using the χ2 and Fisher's exact tests. Logistic regression analysis was used to identify factors influencing stroke and death occurrence. The sample size was determined based on previous experience. P<0.05 was considered to indicate a statistically significant difference.

Results

Comparison of baseline characteristics between the two groups

The baseline characteristics of the two groups are summarised in Table I. Statistically significant differences were observed in left atrial internal diameter,and types of anticoagulants (P<0.05). However, no significant differences were revealed in other baseline characteristics.

Table I A summary presenting the clinical baseline data of patients with AF combined with left atrial thrombus.

Table I

A summary presenting the clinical baseline data of patients with AF combined with left atrial thrombus.

Variant	Radiofrequency ablation combined with anticoagulation Group A (n=55)	Anticoagulation alone Group B (n=37)	t/χ2 valuea	P-value
Age, years	65.2±10.5	68.8±9.0	1.694	0.094
Male [n (%)]	33 (60.0)	24 (64.9)	0.222	0.637
High blood pressure [n (%)]	24 (43.6)	24 (64.9)	3.995	0.056
Diabetes [n (%)]	9 (16.4)	13 (35.1)	4.284	0.068
Diagnostic trait [n (%)]			0.487	0.485
Paroxysmal	7 (12.7)	3 (8.1)
Sustainability	48 (87.3)	34 (91.9)
AF >1 year [n (%)]	22 (40.0)	20 (54.1)	1.761	0.185
Left atrial internal diameter >45 mm [n (%)]	19 (34.5)	23 (62.2)	6.799	0.009
Concomitant medications, n (%)			0.481	0.476
Amiodarone	45 (81.8)	34 (91.9)
β-blocker	10 (18.1)	3 (8.1)
CHADS (Point)	2.0 (1,4)	3.0 (1,4)	0.227	0.821
BNP (ng/l)	580.0 (580,580)	580.0 (397,785)	1.172	0.248
Heart and ears open (mm)	20.3±5.1	19.6±3.7	0.718	0.475
Trunnion (mm)	28.0±4.4	28.2±4.1	0.248	0.805
The distal part of the auricle (mm)	16.1±4.9	15.4±3.5	0.751	0.455
The distal part of the opening (mm)	1.38±0.29	1.36±0.26	0.388	0.699
Types of anticoagulants			20.453	<0.001
Traditional anticoagulants	9 (16.4)	23 (62.2)
New anticoagulants	46 (83.6)	14 (37.8)
Time to review ultrasound (months)	3.0 (3,4)	3.0 (2.5,4)	0.048	0.962

[i] ^aFisher's exact tests or χ² tests for categorical variables, and independent samples t-test for continuous variables. AF, atrial fibrillation; BNP, B-type natriuretic peptide.

Effect of left atrial internal diameter and duration of AF on survival

Statistically significant differences in stroke and death rates were observed between Groups A and B (P<0.05). Stratified analysis revealed no difference in stroke occurrence related to left atrial internal diameter between the two groups. However, a significant impact on death rates was noted in Group B (P<0.05, Table II).

Table II The data demonstrate the effect of left atrial internal diameter on the survival of patients with atrial fibrillation combined with left auricular thrombus in both groups.

Table II

The data demonstrate the effect of left atrial internal diameter on the survival of patients with atrial fibrillation combined with left auricular thrombus in both groups.

	Group A (n=55)		Group B (n=37)
Variant	Left atrial caliber ≤45 mm	Left atrial caliber >45 mm	Left atrial caliber ≤45 mm	Left atrial caliber >45 mm	χ2 valuea	P-value
Stroke [n (%)]					6.106	0.013
Clogged	34 (94.4)	2 (5.6)	10 (71.4)	13 (56.5)
Yes	15 (78.9)	4 (21.1)	4 (28.6)	10 (43.5)
	χ2=3.073, P=0.080		χ2=0.822, P=0.365
End up [n (%)]					9.195	0.002
Clogged	30 (93.3)	11 (57.9)	10 (71.4)	10 (43.5)
Yes	6 (16.7)	8 (42.1)	4 (28.6)	13 (56.5)
	χ2=4.241, P=0.039		χ2=2.737, P=0.098

[i] ^aFisher's exact test or χ² tests for categorical variables.

Stratified analysis of Groups A and B indicated that the duration of AF significantly impacted patient survival. Notably, a significant difference was observed between the duration of AF and the occurrence of stroke in Group B, and an impact on the occurrence of death in both Groups (P<0.05, Table III).

Table III Influence of AF duration on survival of patients with AF combined with left auricular thrombus in both groups.

Table III

Influence of AF duration on survival of patients with AF combined with left auricular thrombus in both groups.

	Group A (n=55)		Group B (n=37)
Variant	AF ≤1 year	AF >1 year	AF ≤1 year	AF >1 year	χ2 valuea	P-value
Stroke [n (%)]					15.947	<0.001
Clogged	31 (93.9)	18 (81.8)	16 (94.1)	7 (35.0)
Yes	2 (6.1)	4 (18.2)	1 (5.9)	13 (65.0)
	χ2=1.995, P=0.158		χ2=13.654, P<0.001
End up [n (%)]					15.351	<0.001
Clogged	28 (84.8)	13 (59.1)	14 (82.4)	6 (30.0)
Yes	5 (15.2)	9 (40.9)	3 (17.6)	14 (70.0)
	χ2=4.615, P=0.032		χ2=10.141, P=0.001

[i] ^aFisher's exact test or χ² tests for categorical variables. AF, atrial fibrillation.

Effect of different anticoagulants on the survival of patients

Groups A and B were further stratified based on the type of anticoagulant used. No statistically significant difference was found regarding the impact of different anticoagulants on stroke and death within Group A (P>0.05). However, in Group B, traditional anticoagulants (warfarin) demonstrated higher survival rates and lower stroke rates compared with NOACs (P<0.05, Table IV).

Table IV A summary demonstrating the effects of different anticoagulants on the survival of patients with atrial fibrillation combined with left auricular thrombus in both groups.

Table IV

A summary demonstrating the effects of different anticoagulants on the survival of patients with atrial fibrillation combined with left auricular thrombus in both groups.

	Group A (n=55)		Group B (n=37)
Variant	Traditional anticoagulants	New anticoagulants	Traditional anticoagulants	New anticoagulants	χ2 valuea	P-value
Stroke [n (%)]					1.078	0.299
Clogged	8 (88.9)	41 (89.1)	19 (82.6)	4 (28.6)
Yes	1 (11.1)	5 (10.9)	4 (17.4)	10 (71.4)
	χ2=0.001, P=0.983		χ2=10.804, P<0.001
End up [n (%)]					1.661	0.198
Clogged	6 (66.7)	35 (76.1)	18 (78.3)	2 (14.3)
Yes	3 (33.3)	11 (23.9)	5 (21.7)	12 (85.7)
	χ2=0.352, P=0.553		χ2=14.342, P<0.001

[i] ^aFisher's exact test.

Univariate analysis affecting patient survival

Univariate analysis revealed that age, hypertension, diabetes mellitus, RA, LVEF, CHADS, BNP and proBNP were significant factors affecting stroke in patients with AF and previous thrombus (P<0.05). Nevertheless, no statistical differences were noted for the remaining factors (Table V).

Table V The data presenting the unifactorial analysis of factors affecting stroke in patients with atrial fibrillation combined with left auricular thrombus.

Table V

The data presenting the unifactorial analysis of factors affecting stroke in patients with atrial fibrillation combined with left auricular thrombus.

	Stroke
Variant	Clogged (n=72)	Yes (n=20)	t/χ2 valuea	P-value
Age, years	65.1±10.4	72.3±6.2	3.903	<0.001
Male [n (%)]	47 (65.3)	10 (50.0)	1.550	0.213
High blood pressure [n (%)]	32 (44.4)	16 (80.0)	7.930	0.005
Diabetes [n (%)]	9 (12.5)	13 (65.0)	23.711	<0.001
Radiofrequency ablation [n (%)]	49 (68.1)	6 (30.0)	9.428	0.002
Diagnostic trait [n (%)]			0.020	0.888
Paroxysmal	8 (11.1)	2 (10.0)
Sustainability	64 (88.9)	18 (90.0)
LVEF (%)	54.8±11.7	47.3±10.3	2.590	0.011
LVEDD (mm)	51.4±7.6	50.3±8.4	0.564	0.574
CHADS (Point)	2 (1,3)	3.5 (3,5)	3.542	0.001
BNP (ng/l)	580 (405,580)	794 (769,845.5)	1.997	0.049
proBNP (ng/l)	2324 (1812,2324)	26052 (2324,2880)	2.004	0.048
Heart and ears open (mm)	19.7±4.2	21.1±5.9	1.166	0.247
Trunnion (mm)	28.1±4.4	28.2±3.6	0.049	0.961
The distal part of the auricle (mm)	15.5±4.1	17.3±5.1	1.648	0.103
The distal part of the opening (mm)	1.4±0.3	1.3±0.2	1.038	0.302

[i] ^aFisher's exact tests or χ² tests for categorical variables, and independent samples t-test for continuous variables. LVEF, left ventricular ejection fraction; LVEDD, dilated left ventricular end-diastolic diameter; BNP, B-type natriuretic peptide.

Additionally, the univariate analysis indicated that RA, trunnion and the distal part of the auricle were significant factors influencing death in patients with AF and left auricle thrombus (P<0.05). However, no statistical differences were observed for other factors (Table VI).

Table VI A summary demonstrates the univariate analysis of factors affecting death in patients with atrial fibrillation combined with left auricular thrombus.

Table VI

A summary demonstrates the univariate analysis of factors affecting death in patients with atrial fibrillation combined with left auricular thrombus.

Variant	Existence (n=61)	End up (n=31)	t/χ2 valuea	P-value
Age, years	66.3±10.3	67.3±9.7	0.462	0.646
Male [n (%)]	39 (63.9)	18 (58.1)	0.300	0.584
High blood pressure [n (%)]	29 (47.5)	19 (61.3)	1.557	0.212
Diabetes [n (%)]	12 (19.7)	10 (32.3)	1.789	0.181
Radiofrequency ablation [n (%)]	41 (67.2)	14 (45.2)	4.157	0.041
Diagnostic trait [n(%)]			0.069	0.793
Paroxysmal	7 (11.5)	3 (9.7)
Sustainability	54 (88.5)	28 (90.3)
LVEF (%)	53.1±12.7	53.2±9.8	0.033	0.974
LVEDD (mm)	51.9±8.1	49.7±7.0	1.275	0.206
CHADS (Point)	2 (1,4)	3 (2,4)	1.421	0.159
BNP (ng/l)	580 (580,580)	580 (580,794)	0.122	0.903
proBNP (ng/l)	2324 (1900,2324)	2324 (2324,2677)	1.717	0.094
Heart and ears open (mm)	19.5±4.2	20.9±5.3	1.461	0.148
Trunnion (mm)	27.5±4.4	29.4±3.6	2.093	0.039
The distal part of the auricle (mm)	15.1±4.0	17.3±4.7	2.326	0.022
The distal part of the opening (mm)	1.4±0.3	1.3±0.2	1.337	0.185

[i] ^aFisher's exact tests or χ² tests for categorical variables, and independent samples t-test for continuous variables. LVEF, left ventricular ejection fraction; LVEDD, dilated left ventricular end-diastolic diameter; BNP, B-type natriuretic peptide.

Multifactorial logistic regression analysis affecting patient survival

Multivariate logistic regression analysis included several attributes such as age, sex, hypertension, diabetes mellitus, diagnostic characteristics, duration of AF. Using stroke occurrence as the dependent variable (yes=‘1’, no=‘0’), factors such as LVEF, LVEDD, CHADS, BNP, proBNP, auricular opening, the long axis of the auricle, distal auricle, distal part of the opening, type of anticoagulant and time to repeat ultrasound (months) were analysed through logistic stepwise regression. The regression analysis identified that an AF of >1 year, a left atrial internal diameter >45 mm, a history of hypertension, diabetes mellitus and high BNP levels were significant risk factors for stroke (P<0.05; Table VII and Fig. 1A).

Figure 1 Forest analysis. (A) Forest analysis affecting stroke in patients with AF combined with left auricular thrombus. (B) Forest analysis affecting the survival of patients with AF combined with left atrial thrombus. AF, atrial fibrillation; BNP, B-type natriuretic peptide.

Table VII The data present the logistic regression analysis affecting stroke in patients with AF combined with left auricular thrombus.

Table VII

The data present the logistic regression analysis affecting stroke in patients with AF combined with left auricular thrombus.

	Univariate regression analyses						Multivariate regression analysis
Variables	B	S.E.	Wald	P	OR	95% CI	B	S.E.	Wald	P	B	S.E.
AF >1 year	3.025	1.086	7.755	0.005	20.594	2.45-173.12	1.617	0.536	9.095	0.003	5.039
Left atrial internal diameter >45 mm	2.092	0.926	5.109	0.024	8.101	1.32-49.699	1.634	0.549	8.849	0.003	5.126	1.746-15.047
Hypertension	2.137	1.048	4.159	0.041	8.474	1.087-66.079
Diabetes mellitus	2.628	0.972	7.314	0.007	13.849	2.062-93.036
Coronary heart disease	2.183	0.925	5.571	0.018	8.875	1.448-54.387
BNP (ng/l)	0.001	0.001	2.767	0.096	1.001	1.000-1.003
Constant	-8.691	2.193	15.707				-2.034	0.516	15.563

[i] CI, confidence interval; AF, atrial fibrillation; BNP, B-type natriuretic peptide; CI, confidence interval; OR, odds ratio.

Similarly, multifactorial logistic regression analysis considered factors such as age, sex, hypertension, diabetes mellitus, diagnostic characteristics, duration of AF and left atrial internal diameter. Using death as the dependent variable (yes=‘1’, no=‘0’), LVEF, LVEDD, CHADS, BNP, proBNP, auricular opening, the long axis of the auricle, distal auricle, distal part of the opening, type of anticoagulant and time to repeat ultrasound (months) were analysed through logistic stepwise regression analysis. The logistic stepwise regression analysis revealed that an AF duration >1 year and a left atrial internal diameter >45 mm were significant risk factors for death (P<0.05; Table VII and Fig. 1B).

Discussion

The present study investigated the effects of RA combined with anticoagulant therapy in patients with AF and left atrial appendage thrombosis post-thrombolysis. The findings of the present study demonstrated a significant reduction in the risk of stroke and mortality when RA was combined with anticoagulant therapy compared with anticoagulant therapy alone. Notably, parameters such as left atrial internal diameter and duration of AF significantly impacted mortality risk. Subgroup analyses revealed no significant difference in survival rates between patients treated with NOACs and those treated with warfarin. Multivariate logistic regression identified AF duration exceeding 1 year, left atrial internal diameter >45 mm, history of hypertension, diabetes mellitus and elevated BNP levels as stroke risk factors. Similarly, AF duration exceeding 1 year and left atrial internal diameter >45 mm emerged as significant risk factors for mortality. Therefore, considering RA in conjunction with anticoagulants is advisable for patients with AF and left atrial appendage thrombosis. Particularly, tailored management strategies are necessary for individuals with prolonged AF duration and enlarged left atrial diameter to mitigate the risks of stroke and mortality.

Recently, RA has emerged as the primary approach for preventing AF recurrence. Despite its recognised safety and efficacy over the past decade, RA still carries potential risks of thrombotic events and bleeding. A previous study reported that ~4.5% of patients with AF treated with RA may experience serious complications, including a haemorrhage incidence of ~2.8% and thrombotic events of ~0.94% (6). Hence, the use of anticoagulation therapy during the perioperative and postoperative phases of RA is recommended. Perioperative anticoagulant administration can significantly mitigate the risks of thrombotic events and major bleeding (7). Furthermore, the 2020 ESC guidelines emphasise that patients with AF should receive anticoagulants for a minimum of 2 months following RA, depending on their stroke risk.

However, there are contrasting findings regarding the efficacy of RA for the treatment of AF (8). A meta-analysis of 11 randomised controlled studies revealed that, while RA improved the quality of life in patients with AF compared with antiarrhythmic drugs, no significant advantage was observed in terms of all-cause mortality, stroke and transient ischaemic attack (9). By contrast, Saglietto et al (10) conducted a meta-analysis of nine studies, demonstrating that RA significantly reduced the risk of heart failure-related death, stroke and hospitalisation in patients with AF compared with medication alone. Mansour et al (11) indicated that patients with AF treated with RA had a lower risk of thromboembolic events and cardiovascular hospitalisation compared with those on antiarrhythmic medications. Similarly, Srivatsa et al (12) found that RA was associated with lower mortality, ischaemic stroke and haemorrhagic stroke in patients with AF. A propensity-matched study by Saliba et al (13) utilising a computerised database from the largest health maintenance organisation in Israel, revealed that radiofrequency CA was linked to reduced stroke/TIA and mortality in patients with high CHA2DS2-Vasc scores. Kim et al (4), utilising the Korean National Health Insurance database, reported that RA significantly reduced the risk of ischaemic stroke and major haemorrhage compared with pharmacological treatment alone. Consistent with these findings (13), the present study suggested that combining anticoagulants with RA lowers the risk of stroke and death compared with anticoagulants alone in patients with AF with left atrial appendage thrombosis. This suggests that the ability of RA to address AF at its source may reduce the recurrence of left atrial appendage thrombosis and subsequent stroke-related deaths. Therefore, early consideration of ablation for patients with AF is advisable as delayed treatment may decrease the success rate of anticoagulation.

In a previous study, thrombosis was treated using low molecular heparin bridged with oral anticoagulants until the international normalised ratio reached a therapeutic range of 2.0-3.0, maintained for at least 3 weeks (14). Oral anticoagulants act by inhibiting the hepatic synthesis of certain factors in the coagulation cascade, such as factors II, VII, IX and X. However, sustained effective anticoagulation may not entirely prevent thromboembolic events in patients with permanent (15) AF and generalised left atrial thrombosis (16). A retrospective series of 43 patients with AF and left atrial thrombosis revealed an increased risk of cerebral embolism and/or death was observed, even in patients under effective anticoagulant therapy with phenprocoumon. Moreover, left atrial thrombus persisted in up to 40% of patients receiving anticoagulants (17). Notably, anticoagulant administration appeared to be less effective in resolving large intracardiac thrombi. Thus, careful selection of effective anticoagulants is crucial for treating patients with AF combined with left auricular thrombus. It was previously suggested that NOAC might be associated with a lower incidence of stroke in the broader AF population compared with warfarin due to fewer haemorrhagic strokes and improved survival (18). However, the comprehensive safety and efficacy of NOAC therapy post-cardiac surgery remain underexplored, resulting in low adoption rates. A previous survey conducted by the European Heart Rhythm Association across 16 centres in 14 countries/regions revealed a low and significantly variable overall adoption of NOACs for AF post-cardiac surgery, with 25% of centres exclusively using warfarin (19). Concerns regarding an increased risk of haemorrhagic pericardial effusion were cited as the primary reason for avoiding NOACs. Furthermore, bleeding complications after cardiac surgery may occur during hospitalisation or shortly after discharge. A study documenting the experience with 72 patients (27 on NOAC vs. 45 on warfarin) found no difference in the incidence rates of short-term stroke or bleeding complications based on the type of oral anticoagulation therapy (20). The results indicated a higher survival rate with oral conventional anticoagulant (warfarin) compared with oral NOACs, potentially attributed to complications associated with NOACs. However, oral warfarin alone may still not be recommended for patients with AF and atrial appendage thrombosis.

Patients with AF and left atrial thrombus are at a high risk of stroke and death (15), underscoring the importance of identifying risk factors for early prevention and prognostic improvement. Previous investigations have highlighted stroke risk factors such as hypertension and diabetes mellitus (15). Elevated plasma BNP levels predict the risk of ischaemic stroke within men from the general population (HR=2.38; 95% CI=1.07-5.29). Left atrial diameter (OR: 1.13; 95% CI: 1.07-1.19) is associated with a heightened risk of future AF following ischaemic stroke (21). However, the left atrial diameter assessed by M-mode echocardiography did not predict stroke (22). Furthermore, it was found that AF with a disease duration of >1 year and left atrial internal diameter >45 mm were risk factors for stroke development in patients. Notably, the risk of death was higher in patients with AF lasting >1 year and left atrial internal diameter >45 mm. Enhanced therapeutic management is often warranted for patients with a combination of these risks.

In conclusion, the present study has investigated the efficacy and adverse events associated with combining RA with anticoagulation treatment in patients with AF and previous left auricular thrombosis. The findings of the present study indicated that the combination of RA with anticoagulants for treating AF with left atrial appendage thrombosis post-thrombolysis yields improved outcomes compared with that of anticoagulants alone. Furthermore, factors such as AF duration (>1 year) and left atrial diameter (>45 mm) may influence patient outcomes. By uniquely examining the efficacy and safety of this combined treatment approach, the present study offers valuable insights into the potential benefits of RA in conjunction with anticoagulation therapy over anticoagulation therapy alone. Nevertheless, there are certain limitations to the present study. The retrospective and non-randomised design of the present study may introduce selection bias and hinder the establishment of causal relationships. Despite statistical adjustments, confirmation of these findings through randomised controlled trials (RCTs) is necessary. While the sample size of 92 patients was adequate for the statistical analyses of the present study, its limitation in generalizability underscores the need for larger, multicentre studies to validate the results obtained from the present study. Multicentre studies could provide a more comprehensive understanding of this topic. Despite employing multivariate logistic regression and propensity score matching, residual confounding due to unmeasured variables remains a possibility. Future research will focus on larger, multicentre RCTs to corroborate the efficacy and safety of combining RA with anticoagulation therapy. Additionally, while the therapeutic effect of administered warfarin appeared higher for those receiving only anticoagulants, further investigation is warranted to confirm these results.

Acknowledgements

Not applicable.

Funding

Funding: The study was supported by The Fund of Traditional Chinese Medicine Science and Technology Project of Shandong Province (grant no. 2020M020).

Availability of data and materials

The data generated in the present study are included in the figures and/or tables of this article.

Authors' contributions

YS, XH and CX designed the study, wrote the first draft of the manuscript and conducted the statistical analysis. MZ participated in the initial experimental design. SW performed the data collection and took part in statistical analysis. MQ provided critical input into the data analysis and interpretation of the results. XL and YD participated in conception and design of the study and revised it critically for important intellectual content. YS, XH and CX confirm the authenticity of all the raw data. All authors read and approved the final version of the manuscript.

Ethics approval and consent to participate

The present study was approved (approval no. IS22031) by the Research Ethics Committee at Shanghai Jiao Tong University Affiliated Chest Hospital (Shanghai, China). All procedures involving human participants conformed to the ethical guidelines of the Declaration of Helsinki. Written informed consent was obtained from all the patients included in the present study.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References