Association between incomplete surgical ligation of left atrial appendage and stroke and systemic embolization Arash Aryana, MS, MD, FHRS,*† Steve K. Singh, MSc, MD,‡ Sheldon M. Singh, MD,§ P. Gearoid O’Neill, MD, FHRS,*† Mark R. Bowers, MD,*† Shelley L. Allen, RN, BSN,* Sammi L. Lewandowski, RN, MSN,* Eleanor C. Vierra, RN, ACNP-C,* André d’Avila, MD, PhD‖ From the *Regional Cardiology Associates, Sacramento, California, †Dignity Health Heart and Vascular Institute, Sacramento, California, ‡Department of Cardiothoracic Surgery, CHI Baylor St. Luke’s Medical Center, Texas Heart Institute, Houston, Texas, §Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada and ‖Instituto de Pesquisa em Arritmia Cardiaca, Hospital Cardiologico– Florianopolis, Florianopolis, Santa Catarina, Brazil. BACKGROUND Surgical exclusion of the left atrial appendage (LAA) can frequently yield incomplete closure. OBJECTIVE We evaluated the ischemic stroke/systemic embolization (SSE) risk in patients with atrial fibrillation (AF) and complete LAA closure (cLAA) vs incompletely surgically ligated LAA (ISLL) and LAA stump after surgical suture ligation. METHODS Seventy-two patients (CHA2DS2-VASc score 4.1 ⫾ 1.9) underwent surgical LAA ligation in conjunction with mitral valve/AF surgery and postoperative LAA evaluation using computerized tomographic angiography.
patient-years of follow-up. The latter risk was nearly 5 times greater than predicted by conventional risk-stratification schemes. Moreover, ISLL emerged as an independent predictor of SSE in univariate analyses and as the sole predictor of SSE in a multivariate analysis. CONCLUSION In patients with AF, ISLL is a predictor of SSE, independent of conventional risk stratification schemes. Consequently, OAC should be strongly considered in this high-risk cohort. KEYWORDS Cardiac surgery; Incomplete closure; Left atrial appendage; Ligation; Stroke
RESULTS Overall, cLAA was detected in 46 of 72 patients (64%), ISLL in 17 patients (24%), and LAA stump in 9 patients (12%). The incidences of oral anticoagulation (OAC) and recurrent AF were similar among the 3 groups during 44 ⫾ 19 months of follow-up. SSE occurred in 2% of patients with cLAA vs 24% with ISLL and 0% with LAA stump (P ¼ .006). None of the patients with SSE were receiving OAC, and all had recurrent AF during follow-up. Additionally, patients with SSE exhibited a significantly smaller ISLL neck diameter (2.8 ⫾ 1.0 vs 7.1 ⫾ 2.1 mm; P ¼ .002). The annualized SSE risk was 1.9% (entire cohort), 6.5% (ISLL patients), 14.4% (ISLL patients not receiving OAC), and 19.0% (ISLL neck diameter r5.0 mm) per 100
ABBREVIATIONS AF ¼ atrial fibrillation; CHADS2 ¼ Congestive heart failure, hypertension, age ≥75 years, diabetes and stroke; CHA2DS2-VASc ¼ Congestive heart failure, hypertension, age ≥75 years, diabetes, stroke, vascular disease, age ≥65 years and sex category; cLAA ¼ complete left atrial appendage closure; ISLL ¼ incompletely surgically ligated left atrial appendage; LAA ¼ left atrial appendage; OAC ¼ oral anticoagulation; SSE ¼ ischemic stroke and systemic embolization
Introduction
thromboembolic risk.3 Incomplete surgical LAA closure can be further classified as incompletely surgically ligated LAA (ISLL) or LAA stump. In this study, we evaluated the risk of ischemic stroke and systemic embolization (SSE) in a contemporary cohort of patients with AF in the absence of rheumatic heart disease, in the setting of ISLL or LAA stump compared with complete LAA closure (cLAA), after surgical suture ligation performed in conjunction with mitral valve or AF surgery.
The left atrial appendage (LAA) is a common site of thrombus formation in patients with atrial fibrillation (AF).1 As such, the practice guidelines recommend surgical LAA closure in patients with AF undergoing mitral valve or Maze surgery.2 However, surgical LAA exclusion, particularly with suture ligation, can often yield incomplete LAA closure,3–6 which may in turn be associated with increased
(Heart Rhythm 2015;12:1431–1437) I 2015 Heart Rhythm Society. All rights reserved.
Methods This study was sponsored by a St. Jude Medical research grant. Address reprint requests and correspondence: Dr Arash Aryana, Regional Cardiology Associates and Dignity Health Heart & Vascular Institute, 3941 J Street, Suite 350, Sacramento, CA 95819. E-mail address: [email protected].
1547-5271/$-see front matter B 2015 Heart Rhythm Society. All rights reserved.
Study patients Patients with nonrheumatic paroxysmal/persistent AF in the absence of mitral stenosis who underwent a first surgical http://dx.doi.org/10.1016/j.hrthm.2015.03.028
1432 LAA suture ligation concurrent with mitral valve or Maze surgery by 5 experienced operators (Mercy General Hospital, Sacramento, California) between January 1, 2008, and December 31, 2012, were enrolled. The surgical approach to LAA ligation consisted of an oversewing technique with a double-layer of running Prolene suture (Ethicon, Inc, Somerville, New Jersey).
Heart Rhythm, Vol 12, No 7, July 2015 electrocardiograms and, whenever present, by use of routine cardiovascular implantable electronic device interrogations. All patients provided informed written consent for their participation in this study. Approval for this study was granted by our institutional review board (Dignity Health Institutional Review Board #14).
Statistical analysis Computerized tomographic angiography All patients underwent outpatient gated cardiac computerized tomographic (CT) angiography Z3 months after surgery to evaluate the status of their LAA. Elective CT angiography for the purpose of this study was avoided within the first 3 months to allow sufficient recovery from the index surgery and to allow for delayed dehiscence of LAA suture ligation, but there was no “blinding period” with respect to detection of SSE in the postoperative follow-up period. Meanwhile, any opacification of the “residual” LAA (partial or complete) was classified as incomplete closure, whereas cLAA was defined as the absence of contrast flow into the presumed anatomic location of the LAA. Incompletely closed LAAs were further classified as (1) ISLL, characterized by the presence of a narrow, constricted LAA neck connecting the intact/nearly intact LAA to the left atrial cavity, or (2) LAA stump, which was defined by presence of a short residual LAA vestige measuring Z1.0 cm in depth (Figure 1). Whenever present, the ISLL morphologies were further subclassified as “cactus,” “chicken wing,” “windsock,” or “cauliflower” based on previously reported LAA morphological criteria.7 Finally, the overall dimensions and volumes of ISLLs and LAA stumps were also recorded. All examinations were carefully reviewed and interpreted by 2 designated, experienced radiologists blinded to the patients’ clinical histories and outcomes. Presence of recurrent AF after LAA ligation was also assessed by use of routine, in-office
Baseline patient demographics and procedural and clinical outcomes were compared among the groups. Continuous variables were analyzed with the 2-sample Student t test or Mann-Whitney test for parametric and nonparametric variables, respectively. The χ2 or Fisher exact test was used for parametric or nonparametric categorical variables, respectively. Predictors of SSE as the primary outcome were discerned by univariable and multivariable regression modeling. The latter included all significant univariate predictors as well as clinically relevant variables, eliminating those that were correlated. This resulted in a parsimonious model after stepwise regression, providing odds ratios and 95% confidence intervals for statistically significant independent predictors of the primary outcome. Time-to-event analysis was performed via Kaplan-Meier curves, compared with logrank and Wilcoxon tests. Additionally, the annualized risk of SSE was calculated for the entire cohort and various subgroups. Late follow-up was complete for 100% of patients at a mean of 44 ⫾ 19 months. For all analyses, P values were 2-sided, and P o .05 was considered significant. Analyses were conducted with SPSS version 20 (IBM SPSS Statistics, Chicago, Illinois).
Results Altogether, 72 patients participated in this study. CT angiography detected cLAA in 46 of 72 patients (64%),
Figure 1 Computed tomography (CT) illustrations of complete left atrial appendage (LAA) closure (cLAA) vs incompletely surgically ligated LAA (ISLL) and LAA stump. Left atrial images segmented from cardiac CT angiograms illustrating cLAA (A), ISLL (B), and LAA stump (C) after prior surgical suture ligation at the time of cardiac surgery. In A, the arrow points to the presumed anatomic location of cLAA. LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein; RIPV ¼ right inferior pulmonary vein.
Aryana et al Table 1
Incomplete LAA Ligation and Stroke Risk
1433
Baseline demographics of patients with cLAA, ISLL, and LAA stump
Characteristic
cLAA (n ¼ 46)
ISLL (n ¼ 17)
LAA stump (n ¼ 9)
P value
Age, y Male Body mass index, Kg/m2 Hypertension Diabetes mellitus Chronic obstructive pulmonary disease Chronic kidney disease Coronary artery disease Congestive heart failure Cardiovascular implantable electronic device Left ventricular ejection fraction, % Left atrial diameter, mm CHADS2 score CHA2DS2-VASc score Paroxysmal AF Atrial flutter SSE
72 ⫾ 10 27 (59) 28 ⫾ 9 33 (72) 9 (20) 14 (30) 15 (33) 13 (28) 28 (61) 10 (22) 51 ⫾ 10 46 ⫾ 8 2.1 ⫾ 0.3 4.0 ⫾ 1.9 24 (52) 9 (20) 1 (2)
71 ⫾ 11 12 (71) 29 ⫾ 5 16 (94) 5 (29) 4 (24) 4 (24) 9 (53) 6 (35) 5 (29) 53 ⫾ 9 47 ⫾ 8 2.1 ⫾ 1.0 4.1 ⫾ 1.9 11 (65) 5 (29) 0 (0)
77 ⫾ 9 6 (67) 28 ⫾ 4 7 (78) 2 (22) 3 (33) 5 (56) 6 (67) 3 (33) 4 (44) 56 ⫾ 18 52 ⫾ 6 1.8 ⫾ 1.0 4.3 ⫾ 1.6 3 (33) 3 (33) 0 (0)
.250 .672 .830 .161 .626 .862 .371 .053 .103 .351 .620 .261 .401 .902 .342 .558 .759
Values are n (%) or mean ⫾ SD. AF ¼ atrial fibrillation; CHADS2 ¼ risk score based on assignment of 1 point each for congestive heart failure, hypertension, age Z75 years, and diabetes mellitus and 2 points for stroke/transient ischemic attack/thromboembolism; CHA2DS2-VASc ¼ risk score based on assignment of 1 point each for congestive heart failure, hypertension, diabetes mellitus, vascular disease, and sex category (1 point for female sex), 1 point for age 65 to 74 years or 2 points for age Z75 years, and 2 points for stroke/transient ischemic attack/thromboembolism; cLAA ¼ complete left atrial appendage closure; ISLL ¼ incompletely surgically ligated left atrial appendage; LAA ¼ left atrial appendage; SSE ¼ ischemic stroke/systemic embolization.
ISLL in 17 patients (24%), and LAA stump in 9 patients (12%). The incidences of ISLL and LAA stump were similar across all operators (P ¼ .273). Baseline patient demographics were similar among the 3 groups (Table 1). This included CHADS2 (risk score based on assignment of 1 point each for congestive heart failure, hypertension, age Z75 years, and diabetes mellitus and 2 points for stroke/transient ischemic attack/thromboembolism) and CHA2DS2-VASc (risk score based on assignment of 1 point each for congestive heart failure, hypertension, diabetes mellitus,
Table 2
vascular disease, and sex category [1 point for female sex], 1 point for age 65 to 74 years or 2 points for age Z75 years, and 2 points for stroke/transient ischemic attack/thromboembolism) risk scores and the incidences of SSE and the relevant risk factors before LAA ligation. During long-term follow-up (Table 2), 53% of patients were diagnosed with recurrent AF. This incidence did not differ among the 3 groups. Although there were no cases of surgery-related SSE, 1 patient with cLAA (2%) developed SSE during follow-up (CHA2DS2-VASc score of 8, 473 days after LAA
Operative and postoperative findings among patients with cLAA, ISLL, and LAA stump during 44 ⫾ 19 months of follow-up
Characteristic Operative Maze surgery Mitral valve surgery Hospital length of stay, d Postoperative Long-term postoperative AF Left atrial diameter, mm Left ventricular ejection fraction, % Pharmacologic therapy Antiarrhythmic drug Beta blocker Statin ACE-I/ARB Diuretic agent Antiplatelet drug OAC SSE Follow-up, mo
cLAA (n ¼ 46)
ISLL (n ¼ 17)
LAA stump (n ¼ 9)
P value
38 (83) 30 (65) 9⫾4
11 (65) 13 (76) 8⫾4
9 (100) 4 (44) 10 ⫾ 3
.084 .132 .613
22 (48) 44 ⫾ 8 54 ⫾ 9
12 (71) 49 ⫾ 7 53 ⫾ 12
4 (44) 49 ⫾ 5 55 ⫾ 12
.251 .059 .660
6 (13) 25 (54) 20 (43) 16 (35) 17 (37) 23 (50) 18 (39) 1 (2) 42 ⫾ 20
4 (24) 6 (35) 12 (71) 8 (47) 10 (59) 10 (59) 10 (59) 4 (24) 44 ⫾ 19
3 (33) 5 (56) 4 (44) 4 (44) 3 (33) 2 (22) 6 (67) 0 (0) 52 ⫾ 17
.166 .415 .173 .677 .275 .200 .181 .006* .220
Values are n (%) or mean ⫾ SD. ACE-I ¼ angiotensin-converting enzyme inhibitor; ARB ¼ angiotensin II receptor blocker; OAC ¼ oral anticoagulant drug. Other abbreviations as in Table 1. * Significant P value.
1434 Table 3
Heart Rhythm, Vol 12, No 7, July 2015 ISLL morphology and LAA dimensions in patients with ISLL and LAA stump and ISLL patients with and without stroke
Parameter ISLL morphology Cactus Chicken wing Windsock Cauliflower LAA dimension Diameter, mm Width, mm Length/depth, mm Volume, cm2 Neck diameter (d), mm
ISLL (n ¼ 17)
LAA stump (n ¼ 9)
P value
ISLL with stroke (n ¼ 4)
ISLL without stroke (n ¼ 13)
P value
6 (35) 6 (35) 4 (24) 1 (6)
… … … …
… … … …
1 (25) 1 (25) 1 (25) 1 (25)
5 (38) 5 (38) 3 (23) 0 (0)
.622 .622 .937 .063
.277 .901 .001* .433 …
12.3 ⫾ 20.0 ⫾ 30.4 ⫾ 8.0 ⫾ 2.8 ⫾
12.7 ⫾ 17.6 ⫾ 32.1 ⫾ 7.6 ⫾ 6.3 ⫾
3.2 2.7 6.6 3.7 2.7
14.6 ⫾ 17.4 ⫾ 22.6 ⫾ 6.3 ⫾ …
4.9 4.4 3.7 3.6
1.5 2.6 3.7 1.2 1.0
12.8 ⫾ 17.0 ⫾ 32.5 ⫾ 7.5 ⫾ 7.4 ⫾
3.6 2.5 7.1 4.1 2.1
.835 .095 .695 .873 .001*
Values are n (%) or mean ⫾ SD. Abbreviations as in Table 1. * Significant P value.
ligation) compared with 4 of 17 patients (24%) with ISLL (CHA2DS2-VASc score of 4.0 ⫾ 1.8, 348 ⫾ 177 days after LAA ligation) and none of the patients with LAA stump (P ¼ .006). The SSE events included 3 ischemic strokes, 1 transient ischemic attack, and 1 systemic cardioembolization to the gastrointestinal tract. Furthermore, the use of antiarrhythmic, antiplatelet, OAC, and other major types of pharmacologic therapies did not differ among the 3 groups after LAA ligation. Although all 5 patients with SSE were diagnosed with recurrent AF post-LAA ligation during follow-up, none were receiving long-term OAC therapy. Aside from SSE, all other operative and postoperative outcomes were similar among the 3 groups. The same was also true when we compared ISLL patients with and without SSE. The ISLL morphologies (eg, cactus, chicken wing, windsock, or cauliflower) and LAA dimensions are illustrated in Table 3. No significant differences in these types of ISLL morphologies were noted in those with and without SSE. Overall LAA dimensions and volume were similar in those with ISLL and LAA stump with the exception of LAA length/depth, which was predictably smaller in patients with LAA stump. Similarly, there were no differences in LAA dimensions and volume when we compared ISLL patients with and without SSE; however, patients with SSE exhibited a significantly smaller ISLL neck diameter than those without SSE (2.8 ⫾ 1.0 vs 7.4 ⫾ 2.1 mm; P ¼ .001). KaplanMeier curves illustrating time-to-event for various groups and subgroups are shown in Figure 2. Overall, the annualized risk of SSE per 100 patient-years of follow-up was determined to be 1.9% for the entire study cohort, 6.5% for ISLL patients, 14.4% for ISLL patients not receiving long-term OAC therapy, and 19.0% for patients with ISLL neck diameter r5.0 mm. In univariate regression analyses, presence of ISLL emerged as an independent predictor of SSE, whereas cLAA, antiplatelet therapy, and CHADS2 or CHA2DS2-VASc risk scores did not (Table 4). Additionally, AF recurrence, lack of OAC therapy, and ISLL neck diameter r5.0 mm were present in all patients with SSE. Hence, all 3 perfectly predicted SSE risk in entirety (100% of the time) with no deviation or 95% confidence
interval applicable. Likewise, in a multivariate analysis, presence of ISLL emerged as a significant independent predictor of SSE risk (odds ratio 20.0, 95% confidence interval 1.92-212.68; P ¼ .012).
Discussion In this study, we evaluated the risk of SSE in a contemporary cohort of patients with AF in the absence of rheumatic heart disease or mitral stenosis, in the setting of ISLL or LAA stump compared with cLAA, after surgical suture ligation performed in association with mitral valve or AF surgery. Several findings in this study are noteworthy. First, the incidence of incomplete LAA closure was 36%, which is similar to that previously reported in the literature.3–6 Second, we observed a higher incidence of SSE in patients with ISLL (24%) than in those with cLAA (2%) or LAA stump (0%). Moreover, this was independent of LAA dimension, volume, or previously described anatomic morphologies (eg, chicken wing vs non-chicken wing). As such, in addition to AF recurrence and lack of OAC, ISLL emerged as an independent predictor of SSE in univariate analyses, and after adjustment for all other variables, it was the only predictor of SSE in a multivariable regression model. Although increased thromboembolic risk in the setting of incomplete LAA closure after surgical ligation has been reported,3 the specific risks associated with various types of incomplete LAA closure have not been investigated previously. In fact, the definition of incomplete LAA closure itself has been nonuniform and somewhat arbitrary across various studies, ranging from “presence of an anatomical structure remaining between the mitral valve base and the left superior pulmonary vein”3,4 to a “residual stump measuring o1.0 cm” or mere presence of “persistent flow into the LAA after surgical exclusion.”5 Yet variations in LAA size and specific morphologies have been shown to correlate with differences in risk for SSE7–10 and LAA thrombus formation.11,12 We will examine several possible explanations that may elucidate this study’s findings. A plausible explanation to account for the significantly higher SSE risk observed among ISLL patients in this study
Aryana et al
Incomplete LAA Ligation and Stroke Risk
1435 Table 4 Significant and nonsignificant predictors of SSE in univariate analyses Variable
OR (95% CI)
P value
Antiplatelet therapy CHADS2 score CHA2DS2-VASc score cLAA ISLL
0.68 2.04 1.27 0.12 16.6
.689 .204 .374 .067 .015*
(0.11–4.38) (0.68–6.16) (0.75–2.15) (0.012–1.16) (1.7–161.4)
OR ¼ odds ratio. * Significant P value.
Figure 2 Kaplan-Meier analyses. Kaplan-Meier curves illustrating the probability of survival free from ischemic stroke/systemic embolization (SSE) in patients with complete left atrial appendage (LAA) closure (cLAA), incompletely surgically ligated LAA (ISLL), and LAA stump (A), patients with ISLL not receiving oral anticoagulant (OAC) therapy (B), and patients with ISLL neck diameter r5 mm (C) as a function of time (days) after surgical LAA ligation.
may have to do with the presence and extent of LAA trabeculation. Extensive LAA trabeculation has been proposed to influence stasis and thrombus formation.7,9,13 By definition, cLAA and LAA stump are generally devoid of extensive trabeculae, whereas the characteristic trabeculation patterns encountered in patients with intact LAAs are also often present in those with ISLL. Another possible explanation may have to do with differences in LAA peak flow velocities in patients with ISLL vs LAA stump. Since reduced peak LAA velocity has been proposed as a risk for LAA thrombus formation,13 one may speculate whether severely reduced peak ISLL velocities could in fact explain the observed difference in SSE risk. However, our unpublished transesophageal echocardiography observations
suggest that although ISLLs may be associated with reduced peak LAA velocities (o20 cm/s), particularly in the setting of AF, peak flow velocities may also be markedly reduced in those with LAA stump. As such, not so much peak flow velocity but anatomic differences appear likely to govern the variations in SSE risk in these different entities. Yet a third explanation may have to do with the LAA neck size. In fact, in this study, we observed the highest SSE risk in those with significantly reduced ISLL neck diameter (r5.0 mm), who exhibited an annualized SSE risk of 19.0% per 100 patientyears of follow-up. Moreover, this also emerged as an independent risk factor for SSE in a univariate analysis. A small LAA neck size has previously been identified as a predictor of thromboembolic stroke in patients with nonligated LAAs.9 Once again, the mechanism for this may be related to a higher degree of stasis within the LAA and its trabeculae. This is also consistent with our observation that patients with an LAA stump, who by definition have a much larger LAA neck diameter, exhibited low SSE risk (no risk in this study). Thus, it is conceivable that the combination of marked LAA neck “stenosis” and preservation of extensive LAA trabeculation as encountered in patients with ISLL may greatly augment the thromboembolic risk in this cohort. Additionally, such a risk appears to be independent of conventional risk-stratification schemes (eg, CHADS2/ CHA2DS2-VASc scores). This is not entirely surprising, because a similar observation is noted, for instance, in patients with valvular AF, in whom the stroke risk remains considerably high ( 17-fold higher) and is also independent of CHADS2 or CHA2DS2-VASc risk scores.1 Overall, these findings suggest that certain anatomic variations in incomplete LAA closure could greatly influence SSE risk. Moreover, these observations also support the notion that manifestation of incomplete LAA closure, primarily through the presence of ISLL, could in fact be worse than no closure at all. As such, the incidence of incomplete LAA closure after surgical ligation may be considered an important quality marker in patients who undergo this procedure, given the potential for undesirable consequences. Presently, screening for incomplete surgical LAA closure is not addressed by the practice guidelines. In addition, most insurance payers in the United States do not recognize or permit surveillance testing. Third, this study also illustrates that the SSE risk associated with ISLL is particularly increased in patients not receiving OAC therapy. The annualized SSE risk among
1436
Heart Rhythm, Vol 12, No 7, July 2015
these patients was 14.4% per 100 patient-years of follow-up, which was nearly 3 times higher than that predicted by conventional risk-stratification schemes (nearly equivalent to a CHADS2 score of 5 or a CHA2DS2-VASc score of 9). In fact, in this study, no SSE events were encountered in patients receiving OAC therapy. Thus, not only does it appear that routine screening for ISLL is necessary, but longterm OAC is rather essential in this cohort, particularly in those with marked ISLL neck stenosis. This gives further credence to the idea that perhaps in some patients with persistently elevated embolic risk and intolerance to longterm OAC therapy, complete ISLL closure by use of alternate strategies may be a reasonable option.14
whereas the other patient was eventually diagnosed with AF as part of the SSE workup. Furthermore, there have also been sporadic reports on the occurrence of LAA thrombus even in patients who remain in sinus rhythm. In a series of consecutive patients with SSE and an absence of significant carotid arterial stenosis, Labovitz and colleagues25 found that 5% of patients in sinus rhythm demonstrated LAA thrombi. Similarly, Vigna and colleagues26 reported an atrial thrombus in 14% of patients with dilated cardiomyopathy who were in sinus rhythm. Therefore, these reports suggest that the LAA may perhaps even serve as a source of thromboembolism in the absence of AF. As such, the optimal approach to management of patients with ISLL without documented AF remains uncertain.
Other clinical implications
Study limitations
Incomplete LAA closure has also been described following various types of endocardial and epicardial percutaneous closure methods.15–19 Moreover, in some cases, incomplete percutaneous LAA closure has been associated with thrombus formation and ischemic stroke.16,17 Bai and colleagues16 reported the presence of a 4-mm peri-Watchman leak in a patient with stroke after LAA closure in whom OAC therapy had been discontinued. However, another study20 evaluating long-term outcomes associated with residual flow after percutaneous LAA closure using the WATCHMAN device found that despite the presence of incomplete LAA closure in 32% of patients at 1 year, residual peridevice leak was not associated with increased embolic risk. In that study, the mean peri-WATCHMAN device leak measured 2.9 ⫾ 1.0 mm (range 1.1-6.0 mm), which is strikingly similar to the ISLL neck diameter (2.8 ⫾ 1.0 mm) in patients who experienced SSE in the current study. In fact, as mentioned, we observed the highest SSE risk among those with the slightest incomplete LAA closure (neck diameter r5.0 mm). Hence, for now, it remains unclear whether incomplete LAA closure after use of certain percutaneous closure techniques behaves differently from that after surgical ligation or whether these discrepant findings somehow reflect variations in embolic risk in diverse patient cohorts. It may be plausible to ask whether a potential benefit derived from cLAA in these studies could be offset in part by the presence of incomplete LAA closure. Certainly, it is interesting that to date, neither surgical nor percutaneous LAA closure methods have demonstrated a superior efficacy compared with OAC therapy.21–23 Ultimately, LAAOS (Left Atrial Appendage Occlusion Study) III, a large, randomized, prospective trial designed to compare the efficacy of LAA closure with excision vs stapled closure to no closure at all,24 may ultimately help address this quandary. Another unresolved clinical dilemma pertains to management of patients with incomplete LAA closure who remain in sinus rhythm. The significantly higher incidence of SSE observed among ISLL patients in this study raises concern regarding the potential for the same result in those who may not be in AF. Also, in the current study, all ISLL patients with SSE except 1 were diagnosed with AF before SSE,
First, this study represents a single-center, nonrandomized analysis of a relatively small cohort of patients with a limited number of events who underwent surgical LAA ligation with concomitant mitral valve or Maze surgery. Second, CT angiography has not been validated as the test of choice for detection of incomplete surgical LAA closure. Although the interpreting radiologists in this study specifically evaluated the CT examinations for presence or absence of this anatomic entity, it is conceivable that incomplete LAA closure could have been missed in certain cases, thereby influencing the reported incidences of ISLL and LAA stump and the relevant findings. Third, the presence or absence of postoperative AF in this study was assessed by routine inoffice electrocardiograms and, whenever present, routine cardiovascular implantable electronic device interrogations. Thus, the incidence of recurrent AF during follow-up could have been underestimated.
Conclusions In patients with nonrheumatic AF who underwent surgical suture ligation of LAA, the presence of ISLL was associated with a significantly higher risk of SSE than cLAA or LAA stump. Although this was independent of the risk determined by the CHADS2 or CHA2DS2-VASc risk scores, ISLL patients not receiving OAC therapy and those with severely constricted ISLL necks exhibited an SSE risk that was 3 to 5 times greater than that predicted by such risk-stratification schemes. As such, routine postoperative ISLL screening and long-term OAC therapy in this high-risk cohort are strongly encouraged.
References 1. Mahajan R, Brooks AG, Sullivan T, Lim HS, Alasady M, Abed HS, Ganesan AN, Nayyar S, Lau DH, Roberts-Thomson KC, Kalman JM, Sanders P. Importance of the underlying substrate in determining thrombus location in atrial fibrillation: implications for left atrial appendage closure. Heart 2012;98:1120–1126. 2. Bonow RO, Carabello BA, Chatterjee K, et al. 2006 Writing Committee Members; American College of Cardiology/American Heart Association Task Force: 2008 Focused update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease): endorsed by the Society of
Aryana et al
3.
4.
5.
6.
7.
8. 9.
10.
11. 12.
13. 14.
15.
Incomplete LAA Ligation and Stroke Risk
Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. Circulation 2008;118:e523–e661. Katz ES, Tsiamtsiouris T, Applebaum RM, Schwartzbard A, Tunick PA, Kronzon I. Surgical left atrial appendage ligation is frequently incomplete: a transesophageal echocardiographic study. J Am Coll Cardiol 2000;36:468–471. García-Fernández MA, Pérez-David E, Quiles J, Peralta J, García-Rojas I, Bermejo J, Moreno M, Silva J. Role of left atrial appendage obliteration in stroke reduction in patients with mitral valve prosthesis: a transesophageal echocardiographic study. J Am Coll Cardiol 2003;42:1253–1258. Healey JS, Crystal E, Lamy A, et al. Left Atrial Appendage Occlusion Study (LAAOS): results of a randomized controlled pilot study of left atrial appendage occlusion during coronary bypass surgery in patients at risk for stroke. Am Heart J 2005;150:288–293. Kanderian AS, Gillinov AM, Pettersson GB, Blackstone E, Klein AL. Success of surgical left atrial appendage closure: assessment by transoesophageal echocardiography. J Am Coll Cardiol 2008;52:924–929. Di Biase L, Santangeli P, Anselmino M, et al. Does the left atrial appendage morphology correlate with the risk of stroke in patients with atrial fibrillation? Results from a multicenter study. J Am Coll Cardiol 2012;60:531–538. Somerville W, Chambers RJ. Systemic embolism in mitral stenosis: relation to the size of the left atrial appendix. Br Med J 1964;2:1167–1169. Khurram IM, Dewire J, Mager M, et al. Relationship between left atrial appendage morphology and stroke in patients with atrial fibrillation. Heart Rhythm 2013;10:1843–1849. Anselmino M, Scaglione M, Di Biase L, Gili S, Santangeli P, Corsinovi L, Pianelli M, Cesarani F, Faletti R, Righi D, Natale A, Gaita F. Left atrial appendage morphology and silent cerebral ischemia in patients with atrial fibrillation. Heart Rhythm 2014;11:2–7. Ernst G, Stöllberger C, Abzieher F, Veit-Dirscherl W, Bonner E, Bibus B, Schneider B, Slany J. Morphology of the left atrial appendage. Anat Rec 1995;242:553–561. Yamamoto M, Seo Y, Kawamatsu N, Sato K, Sugano A, Machino-Ohtsuka T, Kawamura R, Nakajima H, Igarashi M, Sekiguchi Y, Ishizu T, Aonuma K. Complex left atrial appendage morphology and left atrial appendage thrombus formation in patients with atrial fibrillation. Circ Cardiovasc Imaging 2014;7:337–343. Syed FF, DeSimone CV, Friedman PA, Asirvatham SJ. Left atrial appendage exclusion for atrial fibrillation. Cardiol Clin 2014;32:601–625. Aryana A, Cavaco D, Arthur A, O’Neill PG, Dinh H, d’Avila A. Percutaneous endocardial occlusion of incompletely surgically ligated left atrial appendage. J Cardiovasc Electrophysiol 2013;24:968–974. Pillai AM, Kanmanthareddy A, Earnest M, Reddy M, Ferrell R, Nath J, Pillarisetti J, Vallakati A, Lakkireddy D. Initial experience with post Lariat left atrial appendage leak closure with Amplatzer septal occluder device and repeat Lariat application. Heart Rhythm 2014;11:1877–1883.
1437 16. Bai R, Horton RP, Di Biase L, et al. Intraprocedural and long-term incomplete occlusion of the left atrial appendage following placement of the WATCHMAN device: a single center experience. J Cardiovasc Electrophysiol 2012;23:455–461. 17. Lam SC, Bertog S, Sievert H. Incomplete left atrial appendage occlusion and thrombus formation after Watchman implantation treated with anticoagulation followed by further transcatheter closure with a second-generation Amplatzer Cardiac Plug (Amulet device). Catheter Cardiovasc Interv 2015;85:321–327. 18. Price MJ, Gibson DN, Yakubov SJ, et al. Early safety and efficacy of percutaneous left atrial appendage suture ligation: results from the U.S. transcatheter LAA ligation consortium. J Am Coll Cardiol 2014;64:565–572. 19. Miller MA, Gangireddy SR, Doshi SK, Aryana A, Koruth JS, Sennhauser S, d’Avila A, Dukkipati SR, Neuzil P, Reddy VY. Multicenter study on acute and long-term safety and efficacy of percutaneous left atrial appendage closure using an epicardial suture snaring device. Heart Rhythm 2014;11:1853–1859. 20. Viles-Gonzalez JF, Kar S, Douglas P, Dukkipati S, Feldman T, Horton R, Holmes D, Reddy VY. The clinical impact of incomplete left atrial appendage closure with the Watchman Device in patients with atrial fibrillation: a PROTECT AF (Percutaneous Closure of the Left Atrial Appendage Versus Warfarin Therapy for Prevention of Stroke in Patients With Atrial Fibrillation) substudy. J Am Coll Cardiol 2012;59:923–929. 21. Dawson AG, Asopa S, Dunning J. Should patients undergoing cardiac surgery with atrial fibrillation have left atrial appendage exclusion? Interact Cardiovasc Thorac Surg 2010;10:306–311. 22. Holmes DR, Reddy VY, Turi ZG, Doshi SK, Sievert H, Buchbinder M, Mullin CM, Sick P, PROTECT AF Investigators. Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial [published correction appears in Lancet 2009;374:1596]. Lancet 2009;374:534–542. 23. Holmes DR Jr, Kar S, Price MJ, Whisenant B, Sievert H, Doshi SK, Huber K, Reddy VY. Prospective randomized evaluation of the Watchman Left Atrial Appendage Closure device in patients with atrial fibrillation versus long-term warfarin therapy: the PREVAIL trial. J Am Coll Cardiol 2014;64:1–12. 24. Whitlock R, Healey J, Vincent J, et al. Rationale and design of the Left Atrial Appendage Occlusion Study (LAAOS) III. Ann Cardiothorac Surg 2014;3: 45–54. 25. Labovitz AJ for the STEPS Investigators. Transesophageal echocardiography in patients with unexplained cerebral ischemia: multicenter findings. Circulation 1994;90:1–21a. 26. Vigna C, Russo A, De Rito V, Perna G, Villella A, Testa M, Sollazzo V, Fanelli R, Loperfido F. Frequency of left atrial thrombi by transesophageal echocardiography in idiopathic and in ischemic dilated cardiomyopathy. Am J Cardiol 1992;70:1500–1501.
CLINICAL PERSPECTIVES Surgical suture ligation of the left atrial appendage (LAA) is commonly performed in conjunction with mitral valve and Maze surgery; however, this can frequently yield incomplete LAA closure, which in turn can be associated with LAA thrombus formation and thromboembolism. Incomplete surgical LAA closure can be further classified as incompletely surgically ligated LAA (ISLL) or LAA stump. In this study, we evaluated the risk of ischemic stroke and systemic embolization (SSE) in a contemporary cohort of patients with atrial fibrillation in the absence of rheumatic heart disease, with ISLL or LAA stump compared with complete LAA closure. The study found a significantly higher incidence of SSE in patients with ISLL (24%) than in those with complete LAA closure (2%) or LAA stump (0%) during 44 ⫾ 19 months of follow-up (P ¼ .006). Although this appeared to be independent of patient risk as determined by CHADS2 or CHA2DS2VASc scores, ISLL patients not receiving long-term oral anticoagulation and those with severely constricted necks (diameter r5.0 mm) exhibited an SSE risk 3 to 5 times higher than that predicted by such risk-stratification schemes. As such, in addition to atrial fibrillation recurrence and lack of oral anticoagulation, presence of ISLL and ISLL neck diameter r5.0 mm emerged as independent predictors of SSE. Hence, routine screening for ISLL and long-term oral anticoagulation in this high-risk cohort appear to be necessary, particularly in those with marked ISLL neck “stenosis.” It remains unclear whether these findings could also have implications for incomplete LAA closure after percutaneous closure techniques.
patient-years of follow-up. The latter risk was nearly 5 times greater than predicted by conventional risk-stratification schemes. Moreover, ISLL emerged as an independent predictor of SSE in univariate analyses and as the sole predictor of SSE in a multivariate analysis. CONCLUSION In patients with AF, ISLL is a predictor of SSE, independent of conventional risk stratification schemes. Consequently, OAC should be strongly considered in this high-risk cohort. KEYWORDS Cardiac surgery; Incomplete closure; Left atrial appendage; Ligation; Stroke
RESULTS Overall, cLAA was detected in 46 of 72 patients (64%), ISLL in 17 patients (24%), and LAA stump in 9 patients (12%). The incidences of oral anticoagulation (OAC) and recurrent AF were similar among the 3 groups during 44 ⫾ 19 months of follow-up. SSE occurred in 2% of patients with cLAA vs 24% with ISLL and 0% with LAA stump (P ¼ .006). None of the patients with SSE were receiving OAC, and all had recurrent AF during follow-up. Additionally, patients with SSE exhibited a significantly smaller ISLL neck diameter (2.8 ⫾ 1.0 vs 7.1 ⫾ 2.1 mm; P ¼ .002). The annualized SSE risk was 1.9% (entire cohort), 6.5% (ISLL patients), 14.4% (ISLL patients not receiving OAC), and 19.0% (ISLL neck diameter r5.0 mm) per 100
ABBREVIATIONS AF ¼ atrial fibrillation; CHADS2 ¼ Congestive heart failure, hypertension, age ≥75 years, diabetes and stroke; CHA2DS2-VASc ¼ Congestive heart failure, hypertension, age ≥75 years, diabetes, stroke, vascular disease, age ≥65 years and sex category; cLAA ¼ complete left atrial appendage closure; ISLL ¼ incompletely surgically ligated left atrial appendage; LAA ¼ left atrial appendage; OAC ¼ oral anticoagulation; SSE ¼ ischemic stroke and systemic embolization
Introduction
thromboembolic risk.3 Incomplete surgical LAA closure can be further classified as incompletely surgically ligated LAA (ISLL) or LAA stump. In this study, we evaluated the risk of ischemic stroke and systemic embolization (SSE) in a contemporary cohort of patients with AF in the absence of rheumatic heart disease, in the setting of ISLL or LAA stump compared with complete LAA closure (cLAA), after surgical suture ligation performed in conjunction with mitral valve or AF surgery.
The left atrial appendage (LAA) is a common site of thrombus formation in patients with atrial fibrillation (AF).1 As such, the practice guidelines recommend surgical LAA closure in patients with AF undergoing mitral valve or Maze surgery.2 However, surgical LAA exclusion, particularly with suture ligation, can often yield incomplete LAA closure,3–6 which may in turn be associated with increased
(Heart Rhythm 2015;12:1431–1437) I 2015 Heart Rhythm Society. All rights reserved.
Methods This study was sponsored by a St. Jude Medical research grant. Address reprint requests and correspondence: Dr Arash Aryana, Regional Cardiology Associates and Dignity Health Heart & Vascular Institute, 3941 J Street, Suite 350, Sacramento, CA 95819. E-mail address: [email protected].
1547-5271/$-see front matter B 2015 Heart Rhythm Society. All rights reserved.
Study patients Patients with nonrheumatic paroxysmal/persistent AF in the absence of mitral stenosis who underwent a first surgical http://dx.doi.org/10.1016/j.hrthm.2015.03.028
1432 LAA suture ligation concurrent with mitral valve or Maze surgery by 5 experienced operators (Mercy General Hospital, Sacramento, California) between January 1, 2008, and December 31, 2012, were enrolled. The surgical approach to LAA ligation consisted of an oversewing technique with a double-layer of running Prolene suture (Ethicon, Inc, Somerville, New Jersey).
Heart Rhythm, Vol 12, No 7, July 2015 electrocardiograms and, whenever present, by use of routine cardiovascular implantable electronic device interrogations. All patients provided informed written consent for their participation in this study. Approval for this study was granted by our institutional review board (Dignity Health Institutional Review Board #14).
Statistical analysis Computerized tomographic angiography All patients underwent outpatient gated cardiac computerized tomographic (CT) angiography Z3 months after surgery to evaluate the status of their LAA. Elective CT angiography for the purpose of this study was avoided within the first 3 months to allow sufficient recovery from the index surgery and to allow for delayed dehiscence of LAA suture ligation, but there was no “blinding period” with respect to detection of SSE in the postoperative follow-up period. Meanwhile, any opacification of the “residual” LAA (partial or complete) was classified as incomplete closure, whereas cLAA was defined as the absence of contrast flow into the presumed anatomic location of the LAA. Incompletely closed LAAs were further classified as (1) ISLL, characterized by the presence of a narrow, constricted LAA neck connecting the intact/nearly intact LAA to the left atrial cavity, or (2) LAA stump, which was defined by presence of a short residual LAA vestige measuring Z1.0 cm in depth (Figure 1). Whenever present, the ISLL morphologies were further subclassified as “cactus,” “chicken wing,” “windsock,” or “cauliflower” based on previously reported LAA morphological criteria.7 Finally, the overall dimensions and volumes of ISLLs and LAA stumps were also recorded. All examinations were carefully reviewed and interpreted by 2 designated, experienced radiologists blinded to the patients’ clinical histories and outcomes. Presence of recurrent AF after LAA ligation was also assessed by use of routine, in-office
Baseline patient demographics and procedural and clinical outcomes were compared among the groups. Continuous variables were analyzed with the 2-sample Student t test or Mann-Whitney test for parametric and nonparametric variables, respectively. The χ2 or Fisher exact test was used for parametric or nonparametric categorical variables, respectively. Predictors of SSE as the primary outcome were discerned by univariable and multivariable regression modeling. The latter included all significant univariate predictors as well as clinically relevant variables, eliminating those that were correlated. This resulted in a parsimonious model after stepwise regression, providing odds ratios and 95% confidence intervals for statistically significant independent predictors of the primary outcome. Time-to-event analysis was performed via Kaplan-Meier curves, compared with logrank and Wilcoxon tests. Additionally, the annualized risk of SSE was calculated for the entire cohort and various subgroups. Late follow-up was complete for 100% of patients at a mean of 44 ⫾ 19 months. For all analyses, P values were 2-sided, and P o .05 was considered significant. Analyses were conducted with SPSS version 20 (IBM SPSS Statistics, Chicago, Illinois).
Results Altogether, 72 patients participated in this study. CT angiography detected cLAA in 46 of 72 patients (64%),
Figure 1 Computed tomography (CT) illustrations of complete left atrial appendage (LAA) closure (cLAA) vs incompletely surgically ligated LAA (ISLL) and LAA stump. Left atrial images segmented from cardiac CT angiograms illustrating cLAA (A), ISLL (B), and LAA stump (C) after prior surgical suture ligation at the time of cardiac surgery. In A, the arrow points to the presumed anatomic location of cLAA. LIPV ¼ left inferior pulmonary vein; LSPV ¼ left superior pulmonary vein; RIPV ¼ right inferior pulmonary vein.
Aryana et al Table 1
Incomplete LAA Ligation and Stroke Risk
1433
Baseline demographics of patients with cLAA, ISLL, and LAA stump
Characteristic
cLAA (n ¼ 46)
ISLL (n ¼ 17)
LAA stump (n ¼ 9)
P value
Age, y Male Body mass index, Kg/m2 Hypertension Diabetes mellitus Chronic obstructive pulmonary disease Chronic kidney disease Coronary artery disease Congestive heart failure Cardiovascular implantable electronic device Left ventricular ejection fraction, % Left atrial diameter, mm CHADS2 score CHA2DS2-VASc score Paroxysmal AF Atrial flutter SSE
72 ⫾ 10 27 (59) 28 ⫾ 9 33 (72) 9 (20) 14 (30) 15 (33) 13 (28) 28 (61) 10 (22) 51 ⫾ 10 46 ⫾ 8 2.1 ⫾ 0.3 4.0 ⫾ 1.9 24 (52) 9 (20) 1 (2)
71 ⫾ 11 12 (71) 29 ⫾ 5 16 (94) 5 (29) 4 (24) 4 (24) 9 (53) 6 (35) 5 (29) 53 ⫾ 9 47 ⫾ 8 2.1 ⫾ 1.0 4.1 ⫾ 1.9 11 (65) 5 (29) 0 (0)
77 ⫾ 9 6 (67) 28 ⫾ 4 7 (78) 2 (22) 3 (33) 5 (56) 6 (67) 3 (33) 4 (44) 56 ⫾ 18 52 ⫾ 6 1.8 ⫾ 1.0 4.3 ⫾ 1.6 3 (33) 3 (33) 0 (0)
.250 .672 .830 .161 .626 .862 .371 .053 .103 .351 .620 .261 .401 .902 .342 .558 .759
Values are n (%) or mean ⫾ SD. AF ¼ atrial fibrillation; CHADS2 ¼ risk score based on assignment of 1 point each for congestive heart failure, hypertension, age Z75 years, and diabetes mellitus and 2 points for stroke/transient ischemic attack/thromboembolism; CHA2DS2-VASc ¼ risk score based on assignment of 1 point each for congestive heart failure, hypertension, diabetes mellitus, vascular disease, and sex category (1 point for female sex), 1 point for age 65 to 74 years or 2 points for age Z75 years, and 2 points for stroke/transient ischemic attack/thromboembolism; cLAA ¼ complete left atrial appendage closure; ISLL ¼ incompletely surgically ligated left atrial appendage; LAA ¼ left atrial appendage; SSE ¼ ischemic stroke/systemic embolization.
ISLL in 17 patients (24%), and LAA stump in 9 patients (12%). The incidences of ISLL and LAA stump were similar across all operators (P ¼ .273). Baseline patient demographics were similar among the 3 groups (Table 1). This included CHADS2 (risk score based on assignment of 1 point each for congestive heart failure, hypertension, age Z75 years, and diabetes mellitus and 2 points for stroke/transient ischemic attack/thromboembolism) and CHA2DS2-VASc (risk score based on assignment of 1 point each for congestive heart failure, hypertension, diabetes mellitus,
Table 2
vascular disease, and sex category [1 point for female sex], 1 point for age 65 to 74 years or 2 points for age Z75 years, and 2 points for stroke/transient ischemic attack/thromboembolism) risk scores and the incidences of SSE and the relevant risk factors before LAA ligation. During long-term follow-up (Table 2), 53% of patients were diagnosed with recurrent AF. This incidence did not differ among the 3 groups. Although there were no cases of surgery-related SSE, 1 patient with cLAA (2%) developed SSE during follow-up (CHA2DS2-VASc score of 8, 473 days after LAA
Operative and postoperative findings among patients with cLAA, ISLL, and LAA stump during 44 ⫾ 19 months of follow-up
Characteristic Operative Maze surgery Mitral valve surgery Hospital length of stay, d Postoperative Long-term postoperative AF Left atrial diameter, mm Left ventricular ejection fraction, % Pharmacologic therapy Antiarrhythmic drug Beta blocker Statin ACE-I/ARB Diuretic agent Antiplatelet drug OAC SSE Follow-up, mo
cLAA (n ¼ 46)
ISLL (n ¼ 17)
LAA stump (n ¼ 9)
P value
38 (83) 30 (65) 9⫾4
11 (65) 13 (76) 8⫾4
9 (100) 4 (44) 10 ⫾ 3
.084 .132 .613
22 (48) 44 ⫾ 8 54 ⫾ 9
12 (71) 49 ⫾ 7 53 ⫾ 12
4 (44) 49 ⫾ 5 55 ⫾ 12
.251 .059 .660
6 (13) 25 (54) 20 (43) 16 (35) 17 (37) 23 (50) 18 (39) 1 (2) 42 ⫾ 20
4 (24) 6 (35) 12 (71) 8 (47) 10 (59) 10 (59) 10 (59) 4 (24) 44 ⫾ 19
3 (33) 5 (56) 4 (44) 4 (44) 3 (33) 2 (22) 6 (67) 0 (0) 52 ⫾ 17
.166 .415 .173 .677 .275 .200 .181 .006* .220
Values are n (%) or mean ⫾ SD. ACE-I ¼ angiotensin-converting enzyme inhibitor; ARB ¼ angiotensin II receptor blocker; OAC ¼ oral anticoagulant drug. Other abbreviations as in Table 1. * Significant P value.
1434 Table 3
Heart Rhythm, Vol 12, No 7, July 2015 ISLL morphology and LAA dimensions in patients with ISLL and LAA stump and ISLL patients with and without stroke
Parameter ISLL morphology Cactus Chicken wing Windsock Cauliflower LAA dimension Diameter, mm Width, mm Length/depth, mm Volume, cm2 Neck diameter (d), mm
ISLL (n ¼ 17)
LAA stump (n ¼ 9)
P value
ISLL with stroke (n ¼ 4)
ISLL without stroke (n ¼ 13)
P value
6 (35) 6 (35) 4 (24) 1 (6)
… … … …
… … … …
1 (25) 1 (25) 1 (25) 1 (25)
5 (38) 5 (38) 3 (23) 0 (0)
.622 .622 .937 .063
.277 .901 .001* .433 …
12.3 ⫾ 20.0 ⫾ 30.4 ⫾ 8.0 ⫾ 2.8 ⫾
12.7 ⫾ 17.6 ⫾ 32.1 ⫾ 7.6 ⫾ 6.3 ⫾
3.2 2.7 6.6 3.7 2.7
14.6 ⫾ 17.4 ⫾ 22.6 ⫾ 6.3 ⫾ …
4.9 4.4 3.7 3.6
1.5 2.6 3.7 1.2 1.0
12.8 ⫾ 17.0 ⫾ 32.5 ⫾ 7.5 ⫾ 7.4 ⫾
3.6 2.5 7.1 4.1 2.1
.835 .095 .695 .873 .001*
Values are n (%) or mean ⫾ SD. Abbreviations as in Table 1. * Significant P value.
ligation) compared with 4 of 17 patients (24%) with ISLL (CHA2DS2-VASc score of 4.0 ⫾ 1.8, 348 ⫾ 177 days after LAA ligation) and none of the patients with LAA stump (P ¼ .006). The SSE events included 3 ischemic strokes, 1 transient ischemic attack, and 1 systemic cardioembolization to the gastrointestinal tract. Furthermore, the use of antiarrhythmic, antiplatelet, OAC, and other major types of pharmacologic therapies did not differ among the 3 groups after LAA ligation. Although all 5 patients with SSE were diagnosed with recurrent AF post-LAA ligation during follow-up, none were receiving long-term OAC therapy. Aside from SSE, all other operative and postoperative outcomes were similar among the 3 groups. The same was also true when we compared ISLL patients with and without SSE. The ISLL morphologies (eg, cactus, chicken wing, windsock, or cauliflower) and LAA dimensions are illustrated in Table 3. No significant differences in these types of ISLL morphologies were noted in those with and without SSE. Overall LAA dimensions and volume were similar in those with ISLL and LAA stump with the exception of LAA length/depth, which was predictably smaller in patients with LAA stump. Similarly, there were no differences in LAA dimensions and volume when we compared ISLL patients with and without SSE; however, patients with SSE exhibited a significantly smaller ISLL neck diameter than those without SSE (2.8 ⫾ 1.0 vs 7.4 ⫾ 2.1 mm; P ¼ .001). KaplanMeier curves illustrating time-to-event for various groups and subgroups are shown in Figure 2. Overall, the annualized risk of SSE per 100 patient-years of follow-up was determined to be 1.9% for the entire study cohort, 6.5% for ISLL patients, 14.4% for ISLL patients not receiving long-term OAC therapy, and 19.0% for patients with ISLL neck diameter r5.0 mm. In univariate regression analyses, presence of ISLL emerged as an independent predictor of SSE, whereas cLAA, antiplatelet therapy, and CHADS2 or CHA2DS2-VASc risk scores did not (Table 4). Additionally, AF recurrence, lack of OAC therapy, and ISLL neck diameter r5.0 mm were present in all patients with SSE. Hence, all 3 perfectly predicted SSE risk in entirety (100% of the time) with no deviation or 95% confidence
interval applicable. Likewise, in a multivariate analysis, presence of ISLL emerged as a significant independent predictor of SSE risk (odds ratio 20.0, 95% confidence interval 1.92-212.68; P ¼ .012).
Discussion In this study, we evaluated the risk of SSE in a contemporary cohort of patients with AF in the absence of rheumatic heart disease or mitral stenosis, in the setting of ISLL or LAA stump compared with cLAA, after surgical suture ligation performed in association with mitral valve or AF surgery. Several findings in this study are noteworthy. First, the incidence of incomplete LAA closure was 36%, which is similar to that previously reported in the literature.3–6 Second, we observed a higher incidence of SSE in patients with ISLL (24%) than in those with cLAA (2%) or LAA stump (0%). Moreover, this was independent of LAA dimension, volume, or previously described anatomic morphologies (eg, chicken wing vs non-chicken wing). As such, in addition to AF recurrence and lack of OAC, ISLL emerged as an independent predictor of SSE in univariate analyses, and after adjustment for all other variables, it was the only predictor of SSE in a multivariable regression model. Although increased thromboembolic risk in the setting of incomplete LAA closure after surgical ligation has been reported,3 the specific risks associated with various types of incomplete LAA closure have not been investigated previously. In fact, the definition of incomplete LAA closure itself has been nonuniform and somewhat arbitrary across various studies, ranging from “presence of an anatomical structure remaining between the mitral valve base and the left superior pulmonary vein”3,4 to a “residual stump measuring o1.0 cm” or mere presence of “persistent flow into the LAA after surgical exclusion.”5 Yet variations in LAA size and specific morphologies have been shown to correlate with differences in risk for SSE7–10 and LAA thrombus formation.11,12 We will examine several possible explanations that may elucidate this study’s findings. A plausible explanation to account for the significantly higher SSE risk observed among ISLL patients in this study
Aryana et al
Incomplete LAA Ligation and Stroke Risk
1435 Table 4 Significant and nonsignificant predictors of SSE in univariate analyses Variable
OR (95% CI)
P value
Antiplatelet therapy CHADS2 score CHA2DS2-VASc score cLAA ISLL
0.68 2.04 1.27 0.12 16.6
.689 .204 .374 .067 .015*
(0.11–4.38) (0.68–6.16) (0.75–2.15) (0.012–1.16) (1.7–161.4)
OR ¼ odds ratio. * Significant P value.
Figure 2 Kaplan-Meier analyses. Kaplan-Meier curves illustrating the probability of survival free from ischemic stroke/systemic embolization (SSE) in patients with complete left atrial appendage (LAA) closure (cLAA), incompletely surgically ligated LAA (ISLL), and LAA stump (A), patients with ISLL not receiving oral anticoagulant (OAC) therapy (B), and patients with ISLL neck diameter r5 mm (C) as a function of time (days) after surgical LAA ligation.
may have to do with the presence and extent of LAA trabeculation. Extensive LAA trabeculation has been proposed to influence stasis and thrombus formation.7,9,13 By definition, cLAA and LAA stump are generally devoid of extensive trabeculae, whereas the characteristic trabeculation patterns encountered in patients with intact LAAs are also often present in those with ISLL. Another possible explanation may have to do with differences in LAA peak flow velocities in patients with ISLL vs LAA stump. Since reduced peak LAA velocity has been proposed as a risk for LAA thrombus formation,13 one may speculate whether severely reduced peak ISLL velocities could in fact explain the observed difference in SSE risk. However, our unpublished transesophageal echocardiography observations
suggest that although ISLLs may be associated with reduced peak LAA velocities (o20 cm/s), particularly in the setting of AF, peak flow velocities may also be markedly reduced in those with LAA stump. As such, not so much peak flow velocity but anatomic differences appear likely to govern the variations in SSE risk in these different entities. Yet a third explanation may have to do with the LAA neck size. In fact, in this study, we observed the highest SSE risk in those with significantly reduced ISLL neck diameter (r5.0 mm), who exhibited an annualized SSE risk of 19.0% per 100 patientyears of follow-up. Moreover, this also emerged as an independent risk factor for SSE in a univariate analysis. A small LAA neck size has previously been identified as a predictor of thromboembolic stroke in patients with nonligated LAAs.9 Once again, the mechanism for this may be related to a higher degree of stasis within the LAA and its trabeculae. This is also consistent with our observation that patients with an LAA stump, who by definition have a much larger LAA neck diameter, exhibited low SSE risk (no risk in this study). Thus, it is conceivable that the combination of marked LAA neck “stenosis” and preservation of extensive LAA trabeculation as encountered in patients with ISLL may greatly augment the thromboembolic risk in this cohort. Additionally, such a risk appears to be independent of conventional risk-stratification schemes (eg, CHADS2/ CHA2DS2-VASc scores). This is not entirely surprising, because a similar observation is noted, for instance, in patients with valvular AF, in whom the stroke risk remains considerably high ( 17-fold higher) and is also independent of CHADS2 or CHA2DS2-VASc risk scores.1 Overall, these findings suggest that certain anatomic variations in incomplete LAA closure could greatly influence SSE risk. Moreover, these observations also support the notion that manifestation of incomplete LAA closure, primarily through the presence of ISLL, could in fact be worse than no closure at all. As such, the incidence of incomplete LAA closure after surgical ligation may be considered an important quality marker in patients who undergo this procedure, given the potential for undesirable consequences. Presently, screening for incomplete surgical LAA closure is not addressed by the practice guidelines. In addition, most insurance payers in the United States do not recognize or permit surveillance testing. Third, this study also illustrates that the SSE risk associated with ISLL is particularly increased in patients not receiving OAC therapy. The annualized SSE risk among
1436
Heart Rhythm, Vol 12, No 7, July 2015
these patients was 14.4% per 100 patient-years of follow-up, which was nearly 3 times higher than that predicted by conventional risk-stratification schemes (nearly equivalent to a CHADS2 score of 5 or a CHA2DS2-VASc score of 9). In fact, in this study, no SSE events were encountered in patients receiving OAC therapy. Thus, not only does it appear that routine screening for ISLL is necessary, but longterm OAC is rather essential in this cohort, particularly in those with marked ISLL neck stenosis. This gives further credence to the idea that perhaps in some patients with persistently elevated embolic risk and intolerance to longterm OAC therapy, complete ISLL closure by use of alternate strategies may be a reasonable option.14
whereas the other patient was eventually diagnosed with AF as part of the SSE workup. Furthermore, there have also been sporadic reports on the occurrence of LAA thrombus even in patients who remain in sinus rhythm. In a series of consecutive patients with SSE and an absence of significant carotid arterial stenosis, Labovitz and colleagues25 found that 5% of patients in sinus rhythm demonstrated LAA thrombi. Similarly, Vigna and colleagues26 reported an atrial thrombus in 14% of patients with dilated cardiomyopathy who were in sinus rhythm. Therefore, these reports suggest that the LAA may perhaps even serve as a source of thromboembolism in the absence of AF. As such, the optimal approach to management of patients with ISLL without documented AF remains uncertain.
Other clinical implications
Study limitations
Incomplete LAA closure has also been described following various types of endocardial and epicardial percutaneous closure methods.15–19 Moreover, in some cases, incomplete percutaneous LAA closure has been associated with thrombus formation and ischemic stroke.16,17 Bai and colleagues16 reported the presence of a 4-mm peri-Watchman leak in a patient with stroke after LAA closure in whom OAC therapy had been discontinued. However, another study20 evaluating long-term outcomes associated with residual flow after percutaneous LAA closure using the WATCHMAN device found that despite the presence of incomplete LAA closure in 32% of patients at 1 year, residual peridevice leak was not associated with increased embolic risk. In that study, the mean peri-WATCHMAN device leak measured 2.9 ⫾ 1.0 mm (range 1.1-6.0 mm), which is strikingly similar to the ISLL neck diameter (2.8 ⫾ 1.0 mm) in patients who experienced SSE in the current study. In fact, as mentioned, we observed the highest SSE risk among those with the slightest incomplete LAA closure (neck diameter r5.0 mm). Hence, for now, it remains unclear whether incomplete LAA closure after use of certain percutaneous closure techniques behaves differently from that after surgical ligation or whether these discrepant findings somehow reflect variations in embolic risk in diverse patient cohorts. It may be plausible to ask whether a potential benefit derived from cLAA in these studies could be offset in part by the presence of incomplete LAA closure. Certainly, it is interesting that to date, neither surgical nor percutaneous LAA closure methods have demonstrated a superior efficacy compared with OAC therapy.21–23 Ultimately, LAAOS (Left Atrial Appendage Occlusion Study) III, a large, randomized, prospective trial designed to compare the efficacy of LAA closure with excision vs stapled closure to no closure at all,24 may ultimately help address this quandary. Another unresolved clinical dilemma pertains to management of patients with incomplete LAA closure who remain in sinus rhythm. The significantly higher incidence of SSE observed among ISLL patients in this study raises concern regarding the potential for the same result in those who may not be in AF. Also, in the current study, all ISLL patients with SSE except 1 were diagnosed with AF before SSE,
First, this study represents a single-center, nonrandomized analysis of a relatively small cohort of patients with a limited number of events who underwent surgical LAA ligation with concomitant mitral valve or Maze surgery. Second, CT angiography has not been validated as the test of choice for detection of incomplete surgical LAA closure. Although the interpreting radiologists in this study specifically evaluated the CT examinations for presence or absence of this anatomic entity, it is conceivable that incomplete LAA closure could have been missed in certain cases, thereby influencing the reported incidences of ISLL and LAA stump and the relevant findings. Third, the presence or absence of postoperative AF in this study was assessed by routine inoffice electrocardiograms and, whenever present, routine cardiovascular implantable electronic device interrogations. Thus, the incidence of recurrent AF during follow-up could have been underestimated.
Conclusions In patients with nonrheumatic AF who underwent surgical suture ligation of LAA, the presence of ISLL was associated with a significantly higher risk of SSE than cLAA or LAA stump. Although this was independent of the risk determined by the CHADS2 or CHA2DS2-VASc risk scores, ISLL patients not receiving OAC therapy and those with severely constricted ISLL necks exhibited an SSE risk that was 3 to 5 times greater than that predicted by such risk-stratification schemes. As such, routine postoperative ISLL screening and long-term OAC therapy in this high-risk cohort are strongly encouraged.
References 1. Mahajan R, Brooks AG, Sullivan T, Lim HS, Alasady M, Abed HS, Ganesan AN, Nayyar S, Lau DH, Roberts-Thomson KC, Kalman JM, Sanders P. Importance of the underlying substrate in determining thrombus location in atrial fibrillation: implications for left atrial appendage closure. Heart 2012;98:1120–1126. 2. Bonow RO, Carabello BA, Chatterjee K, et al. 2006 Writing Committee Members; American College of Cardiology/American Heart Association Task Force: 2008 Focused update incorporated into the ACC/AHA 2006 guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients With Valvular Heart Disease): endorsed by the Society of
Aryana et al
3.
4.
5.
6.
7.
8. 9.
10.
11. 12.
13. 14.
15.
Incomplete LAA Ligation and Stroke Risk
Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. Circulation 2008;118:e523–e661. Katz ES, Tsiamtsiouris T, Applebaum RM, Schwartzbard A, Tunick PA, Kronzon I. Surgical left atrial appendage ligation is frequently incomplete: a transesophageal echocardiographic study. J Am Coll Cardiol 2000;36:468–471. García-Fernández MA, Pérez-David E, Quiles J, Peralta J, García-Rojas I, Bermejo J, Moreno M, Silva J. Role of left atrial appendage obliteration in stroke reduction in patients with mitral valve prosthesis: a transesophageal echocardiographic study. J Am Coll Cardiol 2003;42:1253–1258. Healey JS, Crystal E, Lamy A, et al. Left Atrial Appendage Occlusion Study (LAAOS): results of a randomized controlled pilot study of left atrial appendage occlusion during coronary bypass surgery in patients at risk for stroke. Am Heart J 2005;150:288–293. Kanderian AS, Gillinov AM, Pettersson GB, Blackstone E, Klein AL. Success of surgical left atrial appendage closure: assessment by transoesophageal echocardiography. J Am Coll Cardiol 2008;52:924–929. Di Biase L, Santangeli P, Anselmino M, et al. Does the left atrial appendage morphology correlate with the risk of stroke in patients with atrial fibrillation? Results from a multicenter study. J Am Coll Cardiol 2012;60:531–538. Somerville W, Chambers RJ. Systemic embolism in mitral stenosis: relation to the size of the left atrial appendix. Br Med J 1964;2:1167–1169. Khurram IM, Dewire J, Mager M, et al. Relationship between left atrial appendage morphology and stroke in patients with atrial fibrillation. Heart Rhythm 2013;10:1843–1849. Anselmino M, Scaglione M, Di Biase L, Gili S, Santangeli P, Corsinovi L, Pianelli M, Cesarani F, Faletti R, Righi D, Natale A, Gaita F. Left atrial appendage morphology and silent cerebral ischemia in patients with atrial fibrillation. Heart Rhythm 2014;11:2–7. Ernst G, Stöllberger C, Abzieher F, Veit-Dirscherl W, Bonner E, Bibus B, Schneider B, Slany J. Morphology of the left atrial appendage. Anat Rec 1995;242:553–561. Yamamoto M, Seo Y, Kawamatsu N, Sato K, Sugano A, Machino-Ohtsuka T, Kawamura R, Nakajima H, Igarashi M, Sekiguchi Y, Ishizu T, Aonuma K. Complex left atrial appendage morphology and left atrial appendage thrombus formation in patients with atrial fibrillation. Circ Cardiovasc Imaging 2014;7:337–343. Syed FF, DeSimone CV, Friedman PA, Asirvatham SJ. Left atrial appendage exclusion for atrial fibrillation. Cardiol Clin 2014;32:601–625. Aryana A, Cavaco D, Arthur A, O’Neill PG, Dinh H, d’Avila A. Percutaneous endocardial occlusion of incompletely surgically ligated left atrial appendage. J Cardiovasc Electrophysiol 2013;24:968–974. Pillai AM, Kanmanthareddy A, Earnest M, Reddy M, Ferrell R, Nath J, Pillarisetti J, Vallakati A, Lakkireddy D. Initial experience with post Lariat left atrial appendage leak closure with Amplatzer septal occluder device and repeat Lariat application. Heart Rhythm 2014;11:1877–1883.
1437 16. Bai R, Horton RP, Di Biase L, et al. Intraprocedural and long-term incomplete occlusion of the left atrial appendage following placement of the WATCHMAN device: a single center experience. J Cardiovasc Electrophysiol 2012;23:455–461. 17. Lam SC, Bertog S, Sievert H. Incomplete left atrial appendage occlusion and thrombus formation after Watchman implantation treated with anticoagulation followed by further transcatheter closure with a second-generation Amplatzer Cardiac Plug (Amulet device). Catheter Cardiovasc Interv 2015;85:321–327. 18. Price MJ, Gibson DN, Yakubov SJ, et al. Early safety and efficacy of percutaneous left atrial appendage suture ligation: results from the U.S. transcatheter LAA ligation consortium. J Am Coll Cardiol 2014;64:565–572. 19. Miller MA, Gangireddy SR, Doshi SK, Aryana A, Koruth JS, Sennhauser S, d’Avila A, Dukkipati SR, Neuzil P, Reddy VY. Multicenter study on acute and long-term safety and efficacy of percutaneous left atrial appendage closure using an epicardial suture snaring device. Heart Rhythm 2014;11:1853–1859. 20. Viles-Gonzalez JF, Kar S, Douglas P, Dukkipati S, Feldman T, Horton R, Holmes D, Reddy VY. The clinical impact of incomplete left atrial appendage closure with the Watchman Device in patients with atrial fibrillation: a PROTECT AF (Percutaneous Closure of the Left Atrial Appendage Versus Warfarin Therapy for Prevention of Stroke in Patients With Atrial Fibrillation) substudy. J Am Coll Cardiol 2012;59:923–929. 21. Dawson AG, Asopa S, Dunning J. Should patients undergoing cardiac surgery with atrial fibrillation have left atrial appendage exclusion? Interact Cardiovasc Thorac Surg 2010;10:306–311. 22. Holmes DR, Reddy VY, Turi ZG, Doshi SK, Sievert H, Buchbinder M, Mullin CM, Sick P, PROTECT AF Investigators. Percutaneous closure of the left atrial appendage versus warfarin therapy for prevention of stroke in patients with atrial fibrillation: a randomised non-inferiority trial [published correction appears in Lancet 2009;374:1596]. Lancet 2009;374:534–542. 23. Holmes DR Jr, Kar S, Price MJ, Whisenant B, Sievert H, Doshi SK, Huber K, Reddy VY. Prospective randomized evaluation of the Watchman Left Atrial Appendage Closure device in patients with atrial fibrillation versus long-term warfarin therapy: the PREVAIL trial. J Am Coll Cardiol 2014;64:1–12. 24. Whitlock R, Healey J, Vincent J, et al. Rationale and design of the Left Atrial Appendage Occlusion Study (LAAOS) III. Ann Cardiothorac Surg 2014;3: 45–54. 25. Labovitz AJ for the STEPS Investigators. Transesophageal echocardiography in patients with unexplained cerebral ischemia: multicenter findings. Circulation 1994;90:1–21a. 26. Vigna C, Russo A, De Rito V, Perna G, Villella A, Testa M, Sollazzo V, Fanelli R, Loperfido F. Frequency of left atrial thrombi by transesophageal echocardiography in idiopathic and in ischemic dilated cardiomyopathy. Am J Cardiol 1992;70:1500–1501.
CLINICAL PERSPECTIVES Surgical suture ligation of the left atrial appendage (LAA) is commonly performed in conjunction with mitral valve and Maze surgery; however, this can frequently yield incomplete LAA closure, which in turn can be associated with LAA thrombus formation and thromboembolism. Incomplete surgical LAA closure can be further classified as incompletely surgically ligated LAA (ISLL) or LAA stump. In this study, we evaluated the risk of ischemic stroke and systemic embolization (SSE) in a contemporary cohort of patients with atrial fibrillation in the absence of rheumatic heart disease, with ISLL or LAA stump compared with complete LAA closure. The study found a significantly higher incidence of SSE in patients with ISLL (24%) than in those with complete LAA closure (2%) or LAA stump (0%) during 44 ⫾ 19 months of follow-up (P ¼ .006). Although this appeared to be independent of patient risk as determined by CHADS2 or CHA2DS2VASc scores, ISLL patients not receiving long-term oral anticoagulation and those with severely constricted necks (diameter r5.0 mm) exhibited an SSE risk 3 to 5 times higher than that predicted by such risk-stratification schemes. As such, in addition to atrial fibrillation recurrence and lack of oral anticoagulation, presence of ISLL and ISLL neck diameter r5.0 mm emerged as independent predictors of SSE. Hence, routine screening for ISLL and long-term oral anticoagulation in this high-risk cohort appear to be necessary, particularly in those with marked ISLL neck “stenosis.” It remains unclear whether these findings could also have implications for incomplete LAA closure after percutaneous closure techniques.
