Abstract
Background:
Atrial fibrillation (AF) is a common cardiac arrhythmia, associated with increased cardiovascular morbidity and mortality including thromboembolic events. The aims of this study were to assess the prevalence of left atrial appendage (LAA) thrombi in Japanese non-valvular atrial fibrillation (NVAF) patients undergoing preprocedural transesophageal echocardiography (TEE) during anticoagulation therapy, and to compare the efficacy of warfarin and direct oral anticoagulants (DOAC).
Methods and Results:
This retrospective study reviewed records of 559 consecutive NVAF patients (445 men; age, 62±11 years) undergoing preprocedural TEE following at least 3 weeks of anticoagulation therapy. Of these, 275 patients had non-paroxysmal AF (49%). LAA thrombus was observed in 15 patients (2.7%). The prevalence of LAA thrombi was similar between the DOAC group (2.6%) and the warfarin group (2.8%, P=0.86). No patients with CHA2DS2-VASc score=0, or paroxysmal AF without prior stroke or transient ischemic attack, had LAA thrombi. On univariate analysis, non-paroxysmal AF, structural heart disease, antiplatelet therapy, larger left atrium, higher brain natriuretic peptide (BNP), reduced LAA flow, and higher CHA2DS2-VASc score were all associated with LAA thrombi. On multivariate analysis, BNP ≥173 pg/mL remained the only independent predictor of LAA thrombi.
Conclusions:
LAA thrombi were found in 2.7% of Japanese NVAF patients scheduled for procedures despite ongoing oral anticoagulation therapy. Incidence of thrombi was similar for patients on DOAC and on warfarin.
Atrial fibrillation (AF) is associated with increased risk of thromboembolic events. Of intracardiac thrombi in AF patients, approximately 90% are located in the left atrial appendage (LAA).1
Transesophageal echocardiography (TEE), with 97% sensitivity and 100% specificity, is considered the gold standard for identification of LAA thrombi and is widely used in clinical practice to screen for the presence of LAA thrombi in AF patients. Several clinical factors have been reported to be associated with LAA thrombi formation, the reported prevalence of which ranges from 0.6 to 18%.1–8
In those previous studies, however, not all patients were on anticoagulants. Recently, it was reported that there was no significant difference in the prevalence of LAA thrombi between those on direct oral anticoagulants (DOAC) and warfarin (4.4% vs. 2.9%, P=0.45), in AF patients undergoing AF ablation and who had been on ≥4 weeks of anticoagulation therapy.9
The aims of the current study were to assess the prevalence of LAA thrombi in Japanese non-valvular atrial fibrillation (NVAF) patients on anticoagulation therapy undergoing preprocedural TEE, and to compare warfarin vs. DOAC with respect to thrombus prevalence.
Methods
Recruitment and TEE Analysis
This single-center retrospective study was conducted by chart review of consecutive Japanese NVAF patients who underwent TEE between July 2011 and October 2015 prior to elective DC cardioversion, cardiac surgery, or catheter ablation of AF. Paroxysmal AF (PAF) was defined as recurrent AF (≥2 episodes) that terminated spontaneously within 7 days.10
Current anticoagulation guidelines for DC cardioversion in patients with AF lasting >48 h recommend the same strategy for warfarin and DOAC, that is, at least 3 weeks of anticoagulation before DC and a minimum of 4 weeks afterwards.10,11
Therefore, patients who had been on anticoagulation therapy for ≥3 weeks prior to the TEE were enrolled. TEE was carried out within 24 h before DC cardioversion. Patient clinical features, echocardiographic parameters, laboratory parameters, medical history (hypertension, diabetes mellitus, chronic heart failure [CHF], vascular disease, cardiomyopathy, prior stroke or transient ischemic attacks [TIA]), and anticoagulant and antiplatelet use just prior to the reference TEE were obtained from the clinical records. The collected data also included brain natriuretic peptide (BNP) and creatinine clearance (Ccr, mL/min, calculated by the Cockcroft-Gault formula) on the day of the TEE, as well as the sequential prothrombin time international normalized ratio (PT-INR; reagent, HemosIL RecombiPlasTin 2G; Instrumentation Laboratory, USA; analyzer, CS-5100, Sysmex, Japan; normal range, 9.0–11.0 s). The CHADS2
and CHA2DS2-VASc score were calculated at the time of the TEE. Only the initial TEE was reviewed for analysis if a patient underwent TEE more than once during the study period.
A retrospective review of TEE reports from medical records was performed. Multiplane TEE was performed using commercially available equipment (Vivid 7; GE Healthcare, Japan) by several cardiologists. All patients gave written consent before TEE. A cine loop of the LA and LAA from 0° to 180° was stored for offline analysis. The pulse-wave Doppler cursor was placed within 1 cm of the LAA orifice for evaluation of the LAA emptying velocity, and 5 consecutive fibrillary emptying waves were averaged. LAA thrombi were defined as round, oval, or irregularly shaped masses that were uniformly echo dense within the LAA, were distinct from the underlying LAA endocardium and the pectinate muscles, and were present in more than 1 imaging plane.12
The patients were followed until April 2016, unless anticoagulants were discontinued or changed, or systemic thromboembolic events occurred. Thromboembolic events included ischemic stroke, TIA, and systemic emboli. Those occurring within 1 week of the medical procedure were excluded because of the possibility of their being procedure-related complications.
Effect of Anticoagulation Therapy
For the patients who were on warfarin, PT-INR data were collected. For patients with 2 or more PT-INR measurements after warfarin was started, the time in therapeutic range (TTR) at the time of the TEE was calculated as the percentage of the period spent in the therapeutic range (PT-INR, 1.6–2.6) of the total duration of warfarin.13,14
The target PT-INR range was set as 1.6–2.6 irrespective of age (≥70 and <70 years), based on the recent Japanese report showing no significant interaction between age and anticoagulation intensity in the effect on thromboembolism and major bleeding.14
For the patients on DOAC, the prescribed dose of drugs was assessed with respect to the recommended dose, in order to identify the patients who were receiving less than the recommended dosage of DOAC. In Japan, the recommended dosage of DOAC is 2-tiered, depending on the patient’s renal function, body weight, age, and concomitant drugs, and are as follows: dabigatran, 150 or 110 mg twice daily; rivaroxaban, 15 or 10 mg once daily; apixaban, 5 or 2.5 mg twice daily; and edoxaban, 60 or 30 mg once daily. Insufficient anticoagulation therapy was defined as TTR <60% or prescription (throughout therapy) of less than the recommended DOAC dose. We found all DOAC prescription doses to have remained constant for all patients.
This study was carried out in accordance with the Declaration of Helsinki. The institutional clinical research ethics committee approved the study protocol. Use of routinely collected anonymous data was approved by the institutional ethics committee and deemed not to require formal patient consent.
Statistical Analysis
Normally distributed continuous variables are expressed as mean±SD, non-normally distributed continuous variables as median (IQR), and as number and percentages for categorical variables. The differences between the continuous variables were assessed using Student’s t-test or Mann-Whitney U-test. Categorical variables were compared using the chi-squared test, or Fisher’s exact test when the data are very unequally distributed among the cells of the table, resulting in expected values <5 in any of the cells of a contingency table. Multivariable logistic regression analysis was performed to identify the independent clinical predictors of LAA thrombi. Statistical significance was defined as P<0.05. SPSS statistics 22 (SPSS, Chicago, IL, USA) was used for all statistical analyses.
Results
A total of 559 consecutive patients (445 men; age, 62±11 years; range, 23–86 years) satisfied the entry criteria. Of these, 275 patients had non-PAF (49%). Patient characteristics are summarized in
Table 1. CHA2DS2-VASc scores of 0–7 were reported in 19.9%, 24.7%, 22.5%, 17.0%, 8.9%, 4.8%, 1.8%, and 0.4% of the patients, respectively. All patients were on anticoagulation therapy for ≥3 weeks, with 311 patients on DOAC (145 on dabigatran, 121 on rivaroxaban, 40 on apixaban, 5 on edoxaban) and 248 on warfarin. A series of INR data was available in 222 patients (90%) on warfarin. TTR varied from 0 to 100% for the whole group (mean, 64±27%); in 90 patients (41%), this was <60%. Among the 311 patients on DOAC, 54 (17%) were on an inappropriately low dose (33 on dabigatran, 18 on rivaroxaban, 1 on apixaban, and 2 on edoxaban).
Table 1.
TEE Subject Characteristics
| |
Total
(n=559) |
LAA thrombi
(n=15) |
No thrombi
(n=544) |
P value |
| Age (years) |
62±11 |
64±13 |
62±11 |
0.67 |
| Male |
445 (79.6) |
9 (60) |
436 (80) |
0.06 |
| Structural heart disease |
29 (5.2) |
4 (26.7) |
27 (4.9) |
0.007 |
| Non-paroxysmal AF |
275 (49.2) |
12 (80) |
263 (48.3) |
0.014 |
Reason for TEE before
ablation/DC/cardiac operation |
502/40/17 |
8/7/0 |
494/33/17 |
0.00001 |
| LAD (mm) |
41±7 |
47±7 |
41±7 |
0.003 |
| LVEF (%) |
62±11 |
53±15 |
62±11 |
0.002 |
| BNP (pg/mL) |
162±384 |
419±474 |
155±379 |
0.008 |
| Ccr (mL/min) |
82±33 |
72±45 |
82±33 |
0.4 |
| LAA flow (m/s) (n=417) |
0.53±0.25 |
0.25±0.09 (n=12) |
0.54±0.25 (n=405) |
0.0001 |
| CHADS2 score |
1.2±1.1 |
2.7±1.2 |
1.1±1.1 |
0.0005 |
| 0 |
173 (30.9) |
0 (0) |
173 (31.8) |
|
| 1 |
221 (39.5) |
4 (6.7) |
217 (39.9) |
|
| ≥2 |
65 (29.6) |
11 (93.3) |
154 (28.3) |
|
| CHA2DS2-VASc score |
1.9±1.5 |
3.7±1.8 |
1.9±1.5 |
0.0001 |
| 0 |
111 (19.9) |
0 (0) |
111 (20.3) |
|
| 1 |
138 (24.7) |
1 (6.7) |
137 (25.3) |
|
| ≥2 |
310 (55.4) |
14 (93.3) |
296 (54.4) |
|
| Anticoagulant |
| Warfarin |
248 (44.4) |
7 (46.7) |
241 (44.3) |
0.86 |
| TTR <60% (n=222) |
90 (40.5) |
3 (42.9) |
87 (40.5) |
|
| DOAC |
311 (55.6) |
8 (53.3) |
303 (55.7) |
0.86 |
| Dabigatran |
145 (25.9) |
3 (20.0) |
142 (26.1) |
|
| Rivaroxaban |
121 (21.6) |
4 (26.7) |
117 (21.5) |
|
| Apixaban |
40 (7.2) |
1 (6.7) |
39 (7.1) |
|
| Edoxaban |
5 (0.9) |
0 (0) |
5 (0.9) |
|
| Inappropriate reduced dose |
54 (17.4) |
1 (12.5) |
53 (17.5) |
|
| Duration of anticoagulation before TEE (days) |
128 (68–348) |
159 (64–1,312) |
145 (66–341) |
0.34 |
| Antiplatelet therapy |
66 (11.8) |
7 (46.7) |
59 (10.8) |
0.0008 |
Data given as mean±SD, median (IQR), or n (%). AF, atrial fibrillation; BNP, brain natriuretic peptide; Ccr, creatinine clearance; DC, DC cardioversion; DOAC, direct oral anticoagulant; LAA, left atrial appendage; LAD, left atrial diameter; LVEF, left ventricular ejection fraction; TEE, transesophageal echocardiography; TTR, time in therapeutic range.
LAA thrombus was observed in 15 patients (2.7%,
Table 2). Among them, 9 patients underwent contrast-enhanced computed tomography (CE-CT) on the same day of TEE, which showed LAA thrombi in 8 patients. Their anticoagulants were dabigatran in 3, rivaroxaban in 4, apixaban in 1, and warfarin in 7 (P=NS,
Figure 1). The prevalence of LAA thrombi was similar between the DOAC and warfarin groups (2.6% in the DOAC group vs. 2.8% in the warfarin group; P=0.86), but patient background was not. The warfarin group had significantly more non-PAF patients (55 vs. 48%, P=0.011), more CHF patients (22 vs. 9%, P=0.0005), and more patients before non-ablation (17 vs. 5%, P=0.0001) compared with the DOAC group. They also had lower Ccr (76 vs. 87 mL/min, P=0.0002) and higher CHA2DS2-VASc score (2.1 vs. 1.8, P=0.03). The prevalence of LAA thrombi increased with higher CHA2DS2-VASc scores (0, 0.7, 1.6, 6.3, 4.0, 3.7, 10, 100% respectively for scores 0–7;
Figure 2), and, in line with this, no patients with CHA2DS2-VASc score=0 had an LAA thrombus.
Table 2.
LAA Thrombi: Patient Clinical Characteristics
Patient
ID no. |
Age
(years) |
Gender |
Type of AF |
Anticoagulation
drug |
PT-INR
≥1.6 (%) |
Inappropriate
dose |
|
| 1 |
82 |
F |
PAF |
Rivaroxaban 10 mg |
– |
No |
|
| 2 |
83 |
F |
Non-PAF |
WF |
35 |
– |
|
| 3 |
75 |
M |
Non-PAF |
WF |
92 |
– |
|
| 4 |
80 |
M |
Non-PAF |
Apixaban 5 mg |
– |
No |
|
| 5 |
52 |
M |
Non-PAF |
WF |
86 |
– |
|
| 6 |
57 |
M |
Non-PAF |
Dabigatran 220 mg |
– |
No |
|
| 7 |
50 |
M |
PAF |
WF |
97 |
– |
|
| 8 |
64 |
F |
Non-PAF |
Rivaroxaban 15 mg |
– |
No |
|
| 9 |
65 |
M |
Non-PAF |
WF |
56 |
– |
|
| 10 |
62 |
M |
PAF |
Dabigatran 300 mg |
– |
No |
|
| 11 |
55 |
M |
Non-PAF |
Rivaroxaban 15 mg |
– |
No |
|
| 12 |
53 |
F |
Non-PAF |
WF |
89 |
– |
|
| 13 |
62 |
F |
Non-PAF |
Rivaroxaban 10 mg |
– |
Yes |
|
| 14 |
41 |
M |
Non-PAF |
Dabigatran 300 mg |
– |
No |
|
| 15 |
73 |
F |
Non-PAF |
WF |
48 |
– |
|
| |
LAD
(mm) |
LVEF
(%) |
BNP
(pg/mL) |
LAA flow
(m/s) |
CHA2DS2-VASc
score |
CHF |
Prior stroke
or TIA |
| 1 |
35 |
78 |
174.2 |
0.20 |
5 |
No |
Yes |
| 2 |
57 |
45 |
307.3 |
0.18 |
7 |
Yes |
No |
| 3 |
48 |
19 |
814.9 |
0.14 |
6 |
Yes |
No |
| 4 |
45 |
75 |
1,930.2 |
No |
4 |
Yes |
No |
| 5 |
47 |
43 |
324.3 |
0.40 |
1 |
Yes |
No |
| 6 |
60 |
51 |
44.7 |
No |
4 |
No |
Yes |
| 7 |
47 |
42 |
669.7 |
0.18 |
3 |
No |
Yes |
| 8 |
43 |
53 |
547 |
0.37 |
3 |
Yes |
No |
| 9 |
47 |
54.6 |
424.5 |
0.32 |
2 |
No |
No |
| 10 |
42 |
70 |
21.1 |
0.23 |
3 |
No |
Yes |
| 11 |
57 |
55 |
174.4 |
0.23 |
3 |
Yes |
Yes |
| 12 |
45 |
58 |
206.2 |
0.20 |
2 |
Yes |
No |
| 13 |
49 |
59 |
192.6 |
0.13 |
3 |
Yes |
No |
| 14 |
40 |
38 |
316.4 |
No |
3 |
Yes |
Yes |
| 15 |
42 |
53 |
136.8 |
0.36 |
7 |
No |
Yes |
| |
Vascular
disease |
Antiplatelet
therapy |
Structural
heart disease |
Switched
drugs |
Outcomes |
Follow-up
(days) |
|
| 1 |
No |
No |
No |
WF |
No follow TEE |
634 |
|
| 2 |
Yes |
Yes (DAPT) |
No |
WF |
Resolution |
1,169 |
|
| 3 |
Yes |
Yes |
No |
WF |
No follow TEE |
690 |
|
| 4 |
No |
No |
Yes |
WF |
CI |
208 |
|
| 5 |
No |
No |
Yes |
WF |
No follow TEE |
764 |
|
| 6 |
No |
No |
No |
WF |
Resolution |
1,271 |
|
| 7 |
No |
Yes |
Yes |
WF |
No follow TEE |
227 |
|
| 8 |
Yes |
No |
No |
WF |
No follow TEE |
583 |
|
| 9 |
No |
No |
Yes |
WF |
Resolution |
1,556 |
|
| 10 |
No |
No |
No |
WF |
Resolution |
1,462 |
|
| 11 |
No |
No |
No |
Dabigatran 300 mg |
Resolution |
846 |
|
| 12 |
No |
Yes |
No |
WF |
Resolution |
943 |
|
| 13 |
No |
Yes |
No |
WF |
LAA amputation |
334 |
|
| 14 |
No |
Yes |
No |
WF |
LAA amputation |
173 |
|
| 15 |
Yes |
Yes |
No |
WF |
Resolution |
409 |
|
CHF, chronic heart failure; CI, cerebral infarction; DAPT, dual antiplatelet therapy; NOAC, non-vitamin K antagonist oral anticoagulants; PAF, paroxysmal atrial fibrillation; PT-INR, prothrombin time international normalized ratio; TIA, transient ischemic attack; WF, warfarin. Other abbreviations as in Table 1.
Among 284 patients with PAF, LAA thrombi was found in 3 patients (1.1%), all of whom had a prior stroke or TIA (Table 2). In the 258 PAF patients who had no prior stroke or TIA, no LAA thrombi was detected among them.
The clinical characteristics of the patients with and without LAA thrombus were compared (Table 1). There were significantly more patients with non-PAF (80 vs. 48%, P=0.014), structural heart disease (27 vs. 5%, P=0.007), and on antiplatelet therapy (47 vs. 11%, P=0.0008) in the LAA thrombi group. The patients with LAA thrombi had a significantly larger left atrium (47 vs. 41 mm, P=0.003), lower left ventricular ejection fraction (LVEF; 53 vs. 62%, P=0.002), higher BNP (419 vs. 155 pg/mL, P=0.008), reduced LAA flow (0.25 vs. 0.54 m/s, P=0.0001), and higher CHA2DS2-VASc score (3.7 vs. 1.9, P=0.0001). Among the conditions included in the CHA2DS2-VASc score, CHF (P=0.00006), prior stroke or TIA (P=0.00009), and vascular disease (P=0.04) were the major drivers of LAA thrombi. There were no significant differences in the number of patients who had insufficient anticoagulation therapy or the duration of the anticoagulation therapy before TEE.
Almost all of the patients who took antiplatelet drugs had risks that were indicated on the univariate analysis for TEE thrombus formation compared with those who did not. That is, the antiplatelet drugs were given to the patients with CHF; hypertension; age ≥65 years; diabetes mellitus; prior stroke or TIA; and vascular disease (P=0.0002, P=0.001, P=0.037, P=0.0002, P=0.0001, and P=0.0001, respectively, compared with those without each risk factor). Prescription of antiplatelet drugs depended on artificial selection; therefore, we did not consider the term in the evaluation of the predictors of LAA thrombus formation on multivariate analysis.
On receiver operating characteristic analysis, left atrial diameter (LAD) >44 mm, BNP >173 pg/mL, or CHA2DS2-VASc score>2 abruptly increased the risk of LAA thrombi formation. No such cut-offs were identified for LVEF or LAA flow. On multivariate analysis, which included structural heart disease, non-PAF, LAD ≥44 mm, BNP ≥173 pg/mL, and CHA2DS2-VASc score≥2 as the variables, BNP ≥173 pg/mL remained an independent predictor of LAA thrombus (OR, 6.10; 95% CI: 1.49–25.4, P=0.012;
Table 3). There were 12 patients (80%) with BNP above this cut-off in the LAA thrombi group, and 114 patients (21%) in the non-LAA thrombi group.
Table 3.
Multivariate Indicators of LAA Thrombi
| Risk factors |
OR |
95% CI |
P-value |
| Structural heart disease |
2.60 |
0.71–9.55 |
0.15 |
| Non-paroxysmal AF |
2.48 |
0.63–9.75 |
0.19 |
| LAD ≥44 mm |
1.35 |
0.41–4.44 |
0.62 |
| BNP ≥173 pg/mL |
6.10 |
1.49–25.4 |
0.013 |
| CHA2DS2-VASc score ≥2 |
5.81 |
0.71–47.4 |
0.10 |
Abbreviations as in Table 1.
The median follow-up duration was 203 days. After LAA thrombus was detected, the 15 patients were given intensive anticoagulant therapy. In those with warfarin, higher PT-INR was set as the target. In those with DOAC, rivaroxaban was switched to dabigatran in 1 patient, and in the other 7 patients DOAC were switched to warfarin. Repeat TEE was performed in 10 of the 15 patients, 7 of whom had resolution of thrombus, while 1 patient on warfarin had a cerebral infarction 7 months after the TEE and 2 others required and successfully underwent LAA amputation (Table 2). Postoperatively, no thrombi were identified in the amputated LAA in the both patients. The 544 patients without LAA thrombi underwent the scheduled procedure after the TEE. Overall, 3 of the patients without LAA thrombus at the time of TEE (0.6%; 1 on warfarin and 2 on rivaroxaban) developed cerebral infarctions during follow-up.
Discussion
In this retrospective, single-center study, the prevalence of LAA thrombus in anti-coagulated Japanese NVAF patients undergoing preprocedural TEE was 2.7%, and this was similar between the warfarin and the DOAC groups. No LAA thrombus was identified in patients with CHA2DS2-VASc score=0, or PAF patients without a prior stroke or TIA. BNP ≥173 pg/mL was the only independent predictor of LAA thrombus.
Prevalence of LAA Thrombi During Anticoagulation
The reported prevalence of LAA thrombi in AF patients is highly variable, ranging from 0.6 to 18%. This may be due to the fact that the proportion of patients on anticoagulation therapy is also highly variable, ranging from 7.9% to 94%, or to variability in the duration of anticoagulation therapy (only a few days in one of the studies), or to the fact that only patients on warfarin were studied.1–8
A well-adjusted warfarin dose reduces the risk of stroke by approximately 60% in AF patients.15
Recently, Frenkel et al assessed the prevalence of LAA thrombi in patients on a ≥4-week course of anticoagulation therapy.9
DOAC included dabigatran, rivaroxaban, and apixaban, and they found no significant difference in the prevalence of LAA thrombi between the patients on warfarin and those on DOAC (2.9% vs. 4.4%),9
similar to the present results. Our study had a greater number of patients, and included edoxaban to cover all 4 currently available DOAC (although the number of patients on edoxaban was 5), and showed that the occurrence of LAA thrombus under current anticoagulation therapy standards is low.
There are some differences in the bleeding complications of DOAC between Asian and Caucasian subjects in worldwide clinical trials.16
Racial differences in the anticoagulation system or the effects of anticoagulation therapy might exist, therefore, optimal anticoagulation for AF patients should consider racial differences.10,14
It is therefore important to obtain data from Japanese patients. To date, Mitamura et al reported that LAA thrombi were observed in 4% of Japanese AF patients on dabigatran.17
In 21% of their patients, dabigatran was used for <3 weeks prior to the TEE. To our knowledge, this is the first report on the prevalence of LAA thrombi in Japanese patients during anticoagulation therapy including all 4 currently available DOAC.
The prevalence of LAA thrombus in the present study was similar between the patients on warfarin and on DOAC, but patient background was not. The warfarin group included more high-risk patients. Given that the clinicians would prescribe warfarin more frequently for patients with higher thromboembolic risks in the present clinical circumstances, it might be impossible to make both groups uniform on retrospective assessment. Nevertheless, this study was a comprehensive survey, reflecting the actual condition. Therefore, although the comparison between warfarin and DOAC could not be judged unconditionally, the similar prevalence of LAA thrombus would have clinical implications.
Although there were a significant proportion of patients with TTR <60% (41%) or taking less than the recommended dose of DOAC (16%), suboptimal anticoagulation therapy was not a predictor of LAA thrombus even on univariate analysis. Such a noteworthy percentage of insufficient anticoagulation therapy might influence the results. Theoretically, the duration and the relevance of the anticoagulation therapy are thought to be related to the thrombus formation, but in the clinical settings neither was significantly different between those with and without LAA thrombus. Although the present NVAF patients had anticoagulation for ≥3 weeks according to the present guidelines, the optimum duration before procedures should be evaluated in a prospective study with more patients.
Predictors of LAA Thrombus
The prevalence of LAA thrombi increases with increasing number of clinical risk factors. LAA emptying velocity, increased LAD, CHADS2
score, non-PAF, impaired LV function, CHF, recent embolic events, diabetes, hypertension, age >75 years, and cardiomyopathy have been reported to be associated with increased LAA thrombi formation.1–4,6,9,12,18
The present findings agree with this, in that LAA thrombi were seen only in those with clinical risk factors, CHA2DS2-VASc score ≥1. Puwanant et al showed that the prevalence of LAA thrombi increased proportionally with increasing CHADS2
score.2
In the present study the relationship between the prevalence of LAA thrombi and CHA2DS2-VASc scores was not entirely linear. The difference might have been due to the effect of the anticoagulation therapy.
In this study, no LAA thrombus was found in patients with CHA2DS2-VASc score=0, or with PAF without a prior stroke or TIA. This suggests that routine TEE prior to invasive procedures would not be required in such low-risk patients.
The use of antiplatelet drugs was significantly more frequent in those with LAA thrombi compared with those without. Antiplatelet therapy was given mainly for ischemic heart disease, or for primary or secondary prevention of thrombotic events. Given that the backgrounds of the patients who took antiplatelet drugs overlapped with the conditions included in the CHA2DS2-VASc score, and the use of antiplatelet drugs depended on artificial selection, we did not include the term in the evaluation of the predictors of LAA thrombus on multivariate analysis. LAA thrombus formation was not prevented in the high-risk patients, even under dual therapy with oral anticoagulants and antiplatelet drugs.
Doukky reported that BNP was an independent predictor of LAA thrombi in NVAF patients.3
None of the Doukky patients with BNP ≤500 pg/mL had LAA thrombi. Similarly, in the present study, on multivariate analysis BNP ≥173 pg/mL was the only independent predictor of LAA thrombi. The BNP cut-off was different between that report and the present study, hence the critical BNP level was also different. Moreover, the characteristics of the patient cohorts were different. The Doukky patients had a higher CHADS2 score than the present ones. It has also been reported that BNP level can predict thromboembolism in patients with NVAF.19
Further study is required to confirm the prognostic value of BNP level on LAA thrombus.
LAA Thrombi and Subsequent Thromboembolic Events
There have been conflicting data pertaining to the association between LAA thrombi and thromboembolic strokes/events. In some previous reports, there was no significant association between LAA thrombi and embolic events in patients with chronic NVAF.2,14
In a TEE substudy of the Stroke Prevention in Atrial Fibrillation-III trial, multivariate analysis of the TEE parameters failed to identify LA thrombus as an independent predictor for embolic events in 192 patients with NVAF.20
In contrast, Stoddard et al recently showed that LA thrombi significantly predicted TIA in 261 patients with AF during a mean follow-up of 30 months.5
The prevalence of LA thrombi in their study was 18%. There have been other reports also showing that LAA thrombi are a precursor of systemic thromboembolism and stroke.10,21,22
There is no consensus for the management of LAA thrombi. In the present study, after having detected LAA thrombi, more rigorous anticoagulant treatment was needed, and LAA amputation as necessary. Despite this, cerebral infarction occurred in 1 of the 15 patients. In contrast, the thrombus resolved in most patients who underwent follow-up TEE. In these clinical settings, warfarin was mostly chosen, setting the target PT-INR considerably higher. Even though thrombi were detected during the anticoagulation therapy, resolution might be obtained with tight and aggressive warfarin control.
Study Limitations
This study included those limitations that are inherent to any retrospective study. First, it was a single-center trial with a limited number of subjects. This study did not include patients with chronic AF in whom catheter ablation or elective DC cardioversion was not attempted. In general, the patients for catheter ablation had lower risk and those for DC cardioversion had higher risk for thromboembolic events. The prevalence of LAA thrombus of 2.7% in this study might depend on the balance between patients for catheter ablation and DC cardioversion. The patient group in this study was distinctly different from that of the AF patient population at large. Therefore, the present conclusions are applicable to this particular cohort and require further validation. Second, TEE was performed by a large number of cardiologists and we used the TEE reports from the medical records, which may have resulted in some inconsistencies in interpretation of whether thrombus was present or not. Although TEE is an established method for detecting LAA thrombi, false positives are still possible. In the present study, the TEE and CE-CT findings were not consistent. Moreover, the postoperative findings in the patients who underwent LAA amputation did not coincide with those of TEE. Third, the anticoagulation therapy depended on each physician in charge. Given that newer DOAC have gradually become available in Japan, the proportion of patients taking each of the 4 DOAC is uneven and skewed towards the older drugs. Fourth, although the effects of oral anticoagulant largely depend on patient compliance, we did not assess this. Fifth, LAA flow data were not available in all patients. We did not evaluate LAA flow depending on the underlying cardiac rhythm, and we may have overlooked the influence of rhythm during TEE on the results. Finally, although BNP (measured just before TEE) ≥173pg/mL was found to be a predictive factor for LAA thrombi, BNP fluctuates. In patients with PAF, BNP may be related to the frequency and duration of the AF attacks and the interval between the last attack and its measurement.
Clinical Implications
AF patients with higher BNP should undergo TEE before clinical procedures even if they have been on oral anticoagulation therapy.
Conclusions
LAA thrombus formation was found in 2.7% of AF patients undergoing current standard oral anticoagulation therapy, and was similar for the patients on DOAC and warfarin. Routine TEE prior to invasive procedures would not be required in low-risk patients, such as those with CHA2DS2-VASc score=0. BNP was a significant predictor of LAA thrombus in patients with NVAF, but further study is required to identify the underlying mechanism.
Acknowledgments
We would like to thank Ms. Hiroko Ozaki for her technical assistance with data collection. This work was supported by a program for women researchers from the Tokyo Medical and Dental University, Japan, through 2015 and 2016.
Funding
There were no sources of funding.
Disclosures
The authors declare no conflict of interest.
References
- 1.
Scherr D, Dalal D, Chilukuri K, Dong J, Spragg D, Henrikson CA, et al. Incidence and predictors of left atrial thrombus prior to catheter ablation of atrial fibrillation. J Cardiovasc Electrophysiol 2009; 20: 379–384.
- 2.
Puwanant S, Varr BC, Shrestha K, Hussain SK, Wilson Tang WH, Gabriel RS, et al. Role of the CHADS2 score in the evaluation of thromboembolic risk in patients with atrial fibrillation undergoing transesophageal echocardiography before pulmonary vein isolation. J Am Coll Cardiol 2009; 54: 2032–2039.
- 3.
Doukky R. B-type natriuretic peptide predicts left atrial appendage thrombus in patients with nonvalvular atrial fibrillation. Echocardiography 2013; 30: 889–895.
- 4.
Rader VJ, Khumri TM, Idupulapati M, Stoner CN, Magalski A, Main ML. Clinical predictors of left atrial thrombus and spontaneous echocardiographic contrast in patients with atrial fibrillation. J Am Soc Echocardiogr 2007; 20: 1181–1185.
- 5.
Stoddard MF, Singh P, Dawn B, Longaker RA. Left atrial thrombus predicts transient ischemic attack in patients with atrial fibrillation. Am Heart J 2003; 145: 676–682.
- 6.
Habara S, Dote K, Kato M, Sasaki S, Goto K, Takemoto H, et al. Prediction of left atrial appendage thrombi in non-valvular atrial fibrillation. Eur Heart J 2007; 28: 2217–2222.
- 7.
Klein AL, Grimm RA, Murray RD, Apperson-Hansen C, Asinger RW, Black IW, et al. Use of transesophageal echocardiography to guide cardioversion in patients with atrial fibrillation. N Engl J Med 2001; 344: 1411–1420.
- 8.
Thambidorai SK, Murray RD, Parakh K, Shah TK, Black IW, Jasper SE, et al. Utility of transesophageal echocardiography in identification of thrombogenic milieu in patients with atrial fibrillation (an ACUTE ancillary study). Am J Cardiol 2005; 96: 935–941.
- 9.
Frenkel D, D’Amato SA, Al-Kazaz M, Markowitz SM, Liu CF, Thomas G, et al. Prevalence of left atrial thrombus detection by transesophageal echocardiography: A comparison of continuous non-vitamin K antagonist oral anticoagulant versus warfarin therapy in patients undergoing catheter ablation for atrial fibrillation. J Am Coll Cardiol 2016; 2: 295–303.
- 10.
Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, et al. 2016 ESC guidelines for the management of atrial fibrillation developed in collaboration with EACTS: The Task Force for the Management of Atrial Fibrillation of the European Society of Cardiology (ESC) developed with the special contribution of the European Heart Rhythm Association (EHRA) of the ESC endorsed by the European Stroke Organisation (ESO). Europace 2016; 18: 1609–1678.
- 11.
Heidbuchel H, Verhamme P, Alings M, Antz M, Diener HC, Hacke W, et al. Updated European Heart Rhythm Association practical guide on the use of non-vitamin-K antagonist anticoagulants in patients with non-valvular atrial fibrillation: Executive summary. Eur Heart J, doi:10.1093/eurheartj/ehw058.
- 12.
Wazni OM, Tsao HM, Chen SA, Chuang HH, Saliba W, Natale A, et al. Cardiovascular imaging in the management of atrial fibrillation. J Am Coll Cardiol 2006; 48: 2077–2084.
- 13.
Rosendaal FR, Cannegieter SC, van der Meer FJ, Briet E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993; 69: 236–239.
- 14.
Yamashita T, Inoue H, Okumura K, Atarashi H, Origasa H; J-Rhythm Registry Investigators. Warfarin anticoagulation intensity in Japanese nonvalvular atrial fibrillation patients: A J-RHYTHM Registry analysis. J Cardiol 2015; 65: 175–177.
- 15.
Hart RG, Pearce LA, Aguilar MI. Meta-analysis: Antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 2007; 146: 857–867.
- 16.
Hori M, Matsumoto M, Tanahashi N, Momomura S, Uchiyama S, Goto S, et al. Rivaroxaban vs. warfarin in Japanese patients with atrial fibrillation: The J-ROCKET AF study. Circ J 2012; 76: 2104–2111.
- 17.
Mitamura H, Nagai T, Watanabe A, Takatsuki S, Okumura K. Left atrial thrombus formation and resolution during dabigatran therapy: A Japanese Heart Rhythm Society report. J Arrhythm 2015; 31: 226–231.
- 18.
Zabalgoitia M, Halperin JL, Pearce LA, Blackshear JL, Asinger RW, Hart RG. Transesophageal echocardiographic correlates of clinical risk of thromboembolism in nonvalvular atrial fibrillation. J Am Coll Cardiol 1998; 31: 1622–1626.
- 19.
Shimizu H, Murakami Y, Inoue S, Ohta Y, Nakamura K, Katoh H, et al. High plasma brain natriuretic polypeptide level as a marker of risk for thromboembolism in patients with nonvalvular atrial fibrillation. Stroke 2002; 33: 1005–1010.
- 20.
The Stroke Prevention in Atrial Fibrillation Investigators Committee on Echocardiography. Transesophageal echocardiographic correlates of thromboembolism in high-risk patients with nonvalvular atrial fibrillation. Ann Intern Med 1998; 128: 639–647.
- 21.
Bayard YL, Omran H, Neuzil P, Thuesen L, Pichler M, Rowland E, et al. PLAATO (Percutaneous Left Atrial Appendage Transcatheter Occlusion) for prevention of cardioembolic stroke in non-anticoagulation eligible atrial fibrillation patients: Results from the European PLAATO study. EuroIntervention 2010; 6: 220–226.
- 22.
Holmes DR Jr, Schwartz RS. Left atrial appendage occlusion eliminates the need for warfarin. Circulation 2009; 120: 1919–1926.
