- Email: [email protected]
Thoracoscopic stapler-and-loop technique for left atrial appendage closure in nonvalvular atrial fibrillation: Mid-term outcomes in 201 patients
Accepted Manuscript Thoracoscopic Stapler-and-loop Technique for Left Atrial Appendage Closure in Nonvalvular Atrial Fibrillation: Mid-term Outcomes in 201 Patients Toshiya Ohtsuka, MD, Takahiro Nonaka, MD, Motoyuki Hisagi, MD, Mikio Ninomiya, MD, Ai Masukawa, MD, Takahiro Ota, MD PII:
S1547-5271(18)30513-7
DOI:
10.1016/j.hrthm.2018.05.026
Reference:
HRTHM 7605
To appear in:
Heart Rhythm
Received Date: 30 March 2018
Please cite this article as: Ohtsuka T, Nonaka T, Hisagi M, Ninomiya M, Masukawa A, Ota T, Thoracoscopic Stapler-and-loop Technique for Left Atrial Appendage Closure in Non-valvular Atrial Fibrillation: Mid-term Outcomes in 201 Patients, Heart Rhythm (2018), doi: 10.1016/j.hrthm.2018.05.026. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1
Thoracoscopic Stapler-and-loop Technique for Left Atrial Appendage Closure in
2
Non-valvular Atrial Fibrillation: Mid-term Outcomes in 201 Patients
4
RI PT
3
Thoracoscopic stapler-and-loop appendage closure
SC
5
Toshiya Ohtsuka, MD,1 Takahiro Nonaka, MD,1 Motoyuki Hisagi, MD,1
7
Mikio Ninomiya, MD,1 Ai Masukawa, MD,2 Takahiro Ota, MD3
8
COI: None
M AN U
6
9
Departments of Cardiovascular Surgery,1 Radiology2 and Neuro-surgery3
11
Tokyo Metropolitan Tama Medical Center, Japan
TE D
10
EP
12
Corresponding author: Toshiya Ohtsuka
14
Department of Cardiovascular Surgery, Tokyo Metropolitan Tama Medical Center
15
2-8-29 Musashi-dai, Fuchu-shi, Tokyo 183-8524 Japan
16
E-mail: [email protected] TEL: +81-42-323-5111 FAX: +81-42-312-9197
AC C
13
17
18
4996 words
1
ACCEPTED MANUSCRIPT
ABSTRACT
19
20
BACKGROUND: Left atrial appendage (LAA) closure can be an alternative to oral
22
anticoagulation to prevent cardiogenic thromboembolisms in non-valvular atrial
23
fibrillation.
24
OBJECTIVE: We retrospectively evaluated safety, completeness and mid-term
25
prevention of our thoracoscopic stapler-and-loop technique for LAA closure.
26
METHODS: Patients operated on between Oct 2008 and Feb 2017 were reviewed.
27
Endoscopic stapler and ligation-loops were employed. Patients received 1-month of
28
anticoagulation before discontinuation. Hospital death and procedure-related major
29
complications (thromboembolisms, hemorrhagic events, phrenic palsy) were the
30
primary composite endpoint for safety, and cardiogenic thromboembolisms were the
31
endpoint for prevention. Brain magnetic resonance imaging investigated new
32
thromboembolic spots 1 year after surgery.
33
RESULTS: There were 201 patients (118 men, 83 women) with a mean age of 74
34
(range 68-94) years, CHA2DS2–Vasc score 4.1 (SD 1.4), HAS-BLED score 2.9 (SD 1.0).
35
The mean operation time was 28 min. ALL LAAs were removed and intra-operative
36
transesophageal echocardiography confirmed completeness of the closure in each
AC C
EP
TE D
M AN U
SC
RI PT
21
2
ACCEPTED MANUSCRIPT
patient. There were no hospital deaths or major procedure-related complications.
38
Follow-up results for 198 patients (98 %) over a mean period of 48 (range 12-110)
39
months revealed that 2 patients developed cardiogenic thromboembolisms (0.25 event
40
per 100 patient-years). Magnetic resonance imaging of 51 patients with a mean
41
CHA2DS2-Vasc score of 4.7 (SD 1.6) revealed one new small spot in each of 2 patients
42
(3.9%, 3.9 spots per 100 patient-years).
43
CONCLUSIONS: Our thoracoscopic stapler-and-loop technique swiftly, safely and
44
completely closed LAAs in patients with non-valvular atrial fibrillation, and provided
45
acceptable mid-term prevention without anticoagulation.
M AN U
SC
RI PT
37
TE D
46
KEY WORDS: non-valvular atrial fibrillation; cardiogenic thromboembolism; left atrial
48
appendage; stapler; thoracoscopic surgery.
50
51
52
AC C
49
EP
47
53
54
3
ACCEPTED MANUSCRIPT
55
INTRODUCTION
56
Following evidence that a transcutaneous left atrial appendage (LAA) closure is
RI PT
57
as protective as warfarin in non-valvular atrial fibrillation (NVAF),1 new surgical
59
procedures have been invented to close LAAs using minimally invasive off-pump
60
techniques. Some evidence of the efficacy of such techniques exists,2-4 but a better
61
understanding of the potential of minimally invasive surgical options would be of benefit
62
to patients looking for alternatives to long-term oral anticoagulation.
M AN U
63
SC
58
We reviewed our 9-year clinical experience of a port-access thoracoscopic stapler-and-loop technique for LAA closure in patients with NVAF to evaluate its safety,
65
completeness and mid-term prevention of cardiogenic thromboembolisms.
66
68
69
METHODS
AC C
67
EP
TE D
64
Our institutional review board approved the clinical protocol for the procedure in
70
2008, and written informed consent to surgery and procedure-related examinations
71
were obtained from each patient.
72
4
ACCEPTED MANUSCRIPT
73
Patient Selection The present technique was offered to patients at risk of cardiogenic
75
thromboembolisms (CHA2DS2-Vasc score ≥ 1) when long-term anticoagulation was
76
deemed difficult (HAS-BLED score ≥ 1) for them, or they had experienced cardiogenic
77
thromboembolisms despite receiving a moderate to high doses of anticoagulants. The
78
procedure was also considered for patients with poor adherence to anticoagulation
79
regimens because, for example, of dementia. Preoperative screening was used to
80
demonstrate that NVAF was the only risk factor for thromboembolisms, and to show that
81
it would be difficult to sustain sinus rhythm with conservative or invasive anti-arrhythmic
82
treatments.
SC
M AN U
TE D
83
RI PT
74
Excluded from the operation were patients deemed unlikely to be able to tolerate 1-month of anticoagulation, those who had a significantly low left-ventricular
85
ejection fraction (< 25 %), those with an inaccessible left thoracic cavity due to
86
pleural/pericardial adhesions, and those unlikely to be able to tolerate a left-lung
87
collapse of longer than 30 min, e.g. patients with terminal pulmonary emphysema. If the
88
LAA contained a clot except in its apical portion, the operation was postponed. Neither
89
the size nor morphology of the LAA limited eligibility for the procedure.
AC C
EP
84
5
ACCEPTED MANUSCRIPT
90
Neurological screening test of the carotid artery was performed; if any lesions were discovered, neurologists followed up and conducted conservative or surgical
92
treatments if necessary.
93
95
Surgical Method (Supplementary Video) and Clinical Pathway
General anesthesia was established through a double-lumen endotracheal
SC
94
RI PT
91
tube. Transesophageal echocardiography (TEE) checked the LAA for clot formation.
97
Four ports were placed in the left lateral thoracic region (Figure 1). No heparin was used
98
during or after surgery. We employed an automatic, bendable cut-and-staple device
99
(ECHELON FLEX™ Powered ENDOPATH® Stapler 60, ETHICON, USA) which can cut
100
and close 6-cm-long tissue with a single cartridge. If any small protrusion was observed
101
at either end of the staple line, an endoscopic ligation tool (ENDOLOOP® Ligature
102
EZ10G, ETHICON, USA) was used to obliterate it. Intraoperative TEE was used for
103
quality control: it guided LAA closure, and recorded the process and results (Figure 2).
105
TE D
EP
AC C
104
M AN U
96
In accordance with our clinical pathway, patients were discharged from hospital
at postoperative day 3.
106
107
6
ACCEPTED MANUSCRIPT
108
109
Anticoagulation Management Anticoagulation was discontinued in accordance with our institutional protocol as follows. Any anticoagulant treatment was suspended 1-2 days before surgery.
111
Anticoagulation (with the same anticoagulant as used preoperatively) was resumed
112
immediately after surgery and continued for 1 month. Enhanced computed tomography
113
was carried out one-month-postoperatively to examine the left atrium, and
114
anticoagulation was thereafter discontinued if the examination proved that the LAA had
115
been successfully removed without a stump (≥ 10 mm in depth) and that there was no
116
clotting along the cut-line. No antiplatelet regimens were used unless patients were
117
already taking them for arteriosclerotic diseases.
SC
M AN U
TE D
119
Endpoints for Procedural Safety
EP
118
RI PT
110
We reviewed hospital charts to identify hospital deaths and major
121
procedure-related complications as the primary composite endpoints for procedural
122
safety, and other complications as secondary ones. Three serious events were defined
123
as major: thromboembolisms, hemorrhagic events and permanent phrenic palsy caused
124
by phrenic nerve injury during pericardiotomy.
AC C
120
125
7
ACCEPTED MANUSCRIPT
126
127
Endpoints for Prevention against Cardiogenic Thromboembolism Neurological investigations were conducted 1, 3, 6 and 12 months after surgery, and yearly thereafter either in the outpatient clinic or by phone or e-mail. Whenever
129
patients had clinical events suggesting ischemia of the brain (disturbance of
130
consciousness, or motor or sensory disorders) or other organs, they were examined
131
and treated. Stroke, transient ischemic attack (TIA) and other ischemic symptoms were
132
judged to be cardiogenic thromboembolisms unless other clear etiologies existed. Using
133
the follow-up data, we identified cardiogenic thromboembolisms and estimated the
134
event rates per 100 patient-years.
137
Brain Magnetic Resonance Imaging Study
At 1-year follow-up, magnetic resonance imaging (MRI) was used to identify
EP
136
TE D
135
M AN U
SC
RI PT
128
new cerebral spots. For this purpose, we enrolled those patients who had experienced
139
stroke or TIA preoperatively and had undergone the diagnostic brain MRI. The
140
preoperative MRI was used as a control and new cerebral spots were counted in the
141
1-year-follow-up images.
AC C
138
142
143
8
ACCEPTED MANUSCRIPT
144
RESULTS
145
147
Patients Characteristics (Table 1)
RI PT
146
Between October 2008 and February 2017, 201 (118 men, 83 women) patients were operated on (3 target patients were excluded). The mean age was 74 years old
149
(range 68-94). NVAF was paroxysmal in 7 patients (4%), persistent in 16 (8%), and
150
long-standing persistent in 178 (88%). 126 patients (63%) had had NVAF for 5 years or
151
longer. The mean CHA2DS2-Vasc and HAS-BLED scores were 4.1 (SD 1.4) and 2.9
152
(SD 1.0), respectively. Nine patients (4%) had been on antiplatelet drugs for
153
arteriosclerotic diseases. Anticoagulants had been used for a mean period of 2.5 years
154
(SD 1.8) [range 0-10] to treat 186 (93%) of the patients (warfarin for 38 of them, and
155
non-vitamin K anticoagulants for 148). Of these 186 patients, 53 (26%) had experienced
156
anticoagulation-associated major bleeding: 5 intracranial, 37 gastrointestinal and 11
157
other hemorrhages. The median left atrial size was 45 mm (range 38-61) on
158
echocardiography. Clot formation had been suspected exclusively at the apical portion
159
of the LAA in 11 patients (5%), although all of them had been continuously
160
anticoagulated. Clots had been positively identified near the ostium or at the mid-portion
161
of the LAA in 2 patients (1%), both of whom had experienced cardiogenic
AC C
EP
TE D
M AN U
SC
148
9
ACCEPTED MANUSCRIPT
thromboembolisms shortly before the examinations; they were therefore treated with
163
warfarin to maintain relatively high INRs for a while (3 months at the longest) until
164
reexamination confirmed the absence of clots; they then underwent surgery.
RI PT
162
165
167
Surgical Outcomes (Table 2)
A single surgeon operated on all patients at a single institute. The approach
SC
166
was port-access thoracoscopy alone in 197 (98%) of the operations, and video-assisted
169
small thoracotomy in the others, because the patients had pericardial or pleural scars.
170
None of the patients underwent cardiopulmonary bypass. The mean operation time for
171
the first half of the patients was 38 min (SD 10) and the second half 28 min (SD 7). The
172
staple-and-loop technique was used to close all LAAs. The completeness of each LAA
173
closure was immediately confirmed with intraoperative TEE. The median cut-length of
174
the LAA was 48 mm (range 35-85 mm).
TE D
EP
There were no hospital deaths and no major procedure-related complications.
AC C
175
M AN U
168
176
Blood loss was less than 50 ml in all operations. Six patients (3.0%) had persistent but
177
minor pulmonary air-leakage, and 1 (0.5%) with dysphasia due to a preoperative stroke
178
suffered aspiration pneumonia. There were 2 re-admissions (1.0%): one patient
179
gradually developed bradycardia, which was treated with a pace-maker implantation,
10
ACCEPTED MANUSCRIPT
and the other developed serious pericarditis, which was treated with steroid
181
administration. 28 patients (14%) suffered post-pericardiotomy syndrome with typical
182
symptoms (chest pain, dry cough, pleural effusions); all symptoms occurred less than 1
183
month after surgery and were abated conservatively with non-steroid anti-inflammatory
184
medication.
SC
RI PT
180
185
187
Follow-up Outcomes
M AN U
186
Computed tomography was carried out 1 month postoperatively in all but 4 (2.0%) of the patients. Three patients (1.5 %) in our early cases revealed a stump
189
(greater than 5 mm but less than 1 cm in depth), and 4 patients (2%) who were operated
190
on before we adopted the endo-loop technique showed a narrow protrusion (shorter
191
than 5 mm) at the caudal end of the staple line. In all of the other patients, the LAA was
192
flat or formed a gentle convexity (Figure 3). A left-atrial clot along the cut line was
193
observed in 2 patients (1.0%); they took anticoagulants for another 2 months, and
194
further examinations revealed 6 months postoperatively that the clots had disappeared.
195
Thus, except for these 2 cases, all of the patients discontinued anticoagulation no later
196
than 1 month after surgery.
AC C
EP
TE D
188
11
ACCEPTED MANUSCRIPT
197
Three patients (1.5%) with CHA2DS2-Vasc scores of 3, 3 and 4, respectively, were lost to follow-up, so outcomes were obtained for 198 patients. The mean and
199
median follow-up periods were 48 (SD 25) [range 12-113] and 46 months. Aggregate
200
patient-years reached 801. In the entire follow-up periods, 5 neurological events
201
occurred, but there were no extra-cerebral thromboembolisms. Three events were
202
proven to be not due to cardiogenic thromboembolisms: 1 patient developed TIA as a
203
stenotic lesion at the carotid artery gradually worsened and therefore underwent an
204
endarterectomy; another patient developed ischemic symptoms due to a known left
205
vertebral artery stenosis and received a bypass operation; and the other patient
206
suffered hypertensive cerebellar bleeding which was treated conservatively. The other 2
207
events were judged to be cardiogenic thromboembolisms (1.0%, 0.25 event per 100
208
patient-years), although satisfactory closure of the LAA had been exhibited in
209
1-month-postoperative enhanced computed tomography. One was TIA (rated modified
210
Rankin score 1), which occurred 2 years after surgery in a 72-year-old woman with
211
long-standing persistent NVAF, a CHA2DS2-Vasc score of 3, and a left atrial size of 45
212
mm. The other event was stroke (rated modified Rankin score 1), which struck 5 months
213
after surgery in a 66-year-old man with long-standing persistent NVAF, a
214
CHA2DS2-Vasc score of 3, and a left atrial size of 50 mm. This was only patient who
AC C
EP
TE D
M AN U
SC
RI PT
198
12
ACCEPTED MANUSCRIPT
215
remained on warfarin maintaining the INR between 2.0 and 2.5 till the end of follow-up.
216
These 5 patients remained free from neurological symptoms thereafter. Four of the patients died (2.0%, 0.53 event per 100 patient-years), 2 of
RI PT
217
malignant or non-cardiac chronic diseases, 1 of chronic heart failure, and 1 of acute
219
coronary syndrome due to a plaque rupture in the left main coronary artery. They were
220
free from cardiogenic thromboembolisms in their follow-up periods.
SC
218
222
Brain-MRI Study Results
M AN U
221
51 patients underwent preoperative brain MRI. The mean age was 75 years
224
(SD 6), with a range of 70-84, and the mean CHA2DS2-Vasc score was 4.7 (SD 1.6),
225
with a range of 2-8. Brain MRI carried out 1 year after surgery revealed 1 new spot in
226
each of only 2 patients (1.0 %, 3.8 spots per 100 patient-years). Each spot was at the
227
cerebral cortex and smaller than 5 mm in diameter.
229
230
EP
AC C
228
TE D
223
231
232
13
ACCEPTED MANUSCRIPT
233
DISCUSSION
234
In 2013, we published a feasibility study of the present technique elsewhere.3
RI PT
235
236
The present report concerns the procedural efficacies: safety, completeness and
237
mid-term prevention of cardiogenic thromboembolisms..
Three kinds of surgical technique for epicardial LAA closure have been
SC
238
previously reported in the literature: suture ligation, stapling, and clipping of the LAA.
240
Suture ligation is the simplest and cheapest method, but the reported results
M AN U
239
were unfavorable in terms of completeness and stroke prevention. Katz et al. found
242
36% of the LAAs were incompletely ligated in concomitant heart surgeries.5 Aryana et al.
243
reported a significantly increased risk (19% per 100 patient-years of follow-up) of stroke
244
or systemic embolization when the neck of the patent LAA was 5 mm or less in
245
diameter.6 No reports of thoracoscopic suture ligation of the LAA have been published
246
since Blackshear et al. reported on 15 patients in 2003,2 so surgeons would probably be
247
well advised to avoid this technique (either via open or thoracoscopic approach) except
248
for the LAAs with a very small, round ostium.
249
250
AC C
EP
TE D
241
Although the endo-stapler is a commonly used and, therefore, reasonably priced device, some surgeons have questioned its performance in LAA closure. Lee et
14
ACCEPTED MANUSCRIPT
al. compared 3 different techniques and found that stapling left unacceptable residual
252
stumps in 60% of their subjects.7 Kanderian et al. reported no success with the device.8
253
One possible reason for such failure may be that few heart surgeons are familiar with
254
endo-staplers, and another likely reason is that LAA closure was conducted via
255
sternotomy: the LAA is deeply located in the pericardial cavity via sternotomy, so it is
256
difficult to direct the device properly along the ostium of the LAA. Besides, viewed from
257
above, its cranial part is often hidden behind the left pulmonary artery, so there is a
258
residual stump there. On the other hand, thoracoscopy allows visualization of all parts of
259
the LAA, and a stapler can be properly introduced to the LAA base through a port in the
260
lateral thoracic region. Ruttkay et al. completely closed a giant LAA with a similar
261
approach to ours.9
SC
M AN U
TE D
Clipping is the most expensive method, but clipping devices are remarkably
EP
262
RI PT
251
attractive because they permit atraumatic and arguably more surgeon-friendly
264
techniques than stapling of the LAA. Caliskan et al. reported their clinical experience in
265
open heart surgery in which all LAAs were clipped with no sign of significant stumps and
266
satisfactory mid-term results in terms of stroke prevention (0.5 per 100 patient-years),10
267
a similar result to ours. Osmancik et al. reported a very high success rate (97.5%) in 40
268
cases of thoracoscopic clip ligation, but they found stumps deeper than 10 mm in 45%
AC C
263
15
ACCEPTED MANUSCRIPT of the patients, and a volume of 3.6 ± 1.9 ml.11 The clinical implications of such residual
270
stumps are unclear, but the quality of closure seems to be compromised by use of the
271
device off pump. Once the clip has been deployed on the LAA, its position can never be
272
adjusted and, unlike with our endo-loop technique, no touch-up measures are available
273
to obliterate below-clip stumps. This kind of problem needs to be overcome by further
274
material improvements.
SC
In the field of LAA management, transcutaneous catheter techniques have also
M AN U
275
RI PT
269
evolved although they are before approval in Japan: device implantation into the LAA
277
with occluders, e.g. the Watchman, and extracardiac ligation with the Lariat device.12-14
278
These techniques are no doubt less invasive (or less incisional) than any surgical
279
methods, but have controversial issues to be discussed in comparison with our method.
280
First, procedural safety is a pivotal point. While Watchman implantation seems
EP
TE D
276
to have become less risky as far as the results of the PREVAIL versus PROTECT AF
282
study are concerned,15 the Lariat technique was shown by Srivasta et al. to have yielded
283
suboptimal outcomes, causing cardiac tamponade in 7.5% of patients, LAA laceration in
284
1.8%, and death in 0.3%.14 Thoracoscopy clearly shows that LAA tissue is thick and
285
strong at the basal part, but very thin and breakable at the other parts; we therefore
AC C
281
16
ACCEPTED MANUSCRIPT
never pinch and pull up the LAA with our technique. Attention needs to be paid to this
287
anatomical characteristic to further improve the safety of the transcutaneous methods.
288
The second issue is the procedural limitations dictated by the LAA’s size and
RI PT
286
morphology. Although our method is not limited according to LAA anatomy, the
290
transcutaneous techniques certainly are. According to the instruction for use, the
291
Watchman cannot be implanted if the maximum ostial width exceeds 31 mm. Bartus et
292
al. recommended the Lariat technique not be used when the LAA is wider than 40 mm,
293
when the LAA is superiorly oriented with its apex behind the pulmonary trunk, and when
294
the lobes in a multi-lobed LAA are oriented in different planes exceeding 40 mm.16 No
295
reports on either technique have clarified what percentage of LAAs are actually
296
excluded because of their size or morphology, although it was clear from the Protect AF
297
study that 1945/4998 cases (39%) were excluded for clinical or echocardiographic
298
reasons.1 Since previous reports show that LAAs with a larger size or more complex
299
shape pose a greater thromboembolic risk to patients,17-20 we believe that our technique
300
should treat LAAs with such risky anatomies, rather than transcutaneous methods.
M AN U
TE D
EP
AC C
301
SC
289
Residual gaps are also an issue. Although we remove the LAA in every patient,
302
after 1 year of follow-up, Pillarisetti et al. reported that the Watchman left a gap in 21%
303
of subject patients, and the Lariat device in 14%.21 Yarlagadda et al. stressed that
17
ACCEPTED MANUSCRIPT
ligation with the Lariat created fewer leaks than surgical ligation, and that the recurrent
305
gaps had no link with cardiogenic thromboembolic events after either technique.22
306
Nevertheless, as Aryana warned,23 the patent LAA is able to generate thrombo-emboli.
307
Plugs or clips might be able to control central leaks after ligation with the Lariat,24,25 but it
308
would be extremely difficult to stop edge leaks after device implantation.
After Watchman implantation, anticoagulation management continues until the
SC
309
RI PT
304
endothelial coverage of the device is completed, although progression varies from
311
patient to patient.1 Our protocol might be simpler: anticoagulants are administered for 1
312
month postoperatively. We established this protocol because procedure-associated
313
inflammatory reactions in the acute post-operative period might cause clotting at the
314
cut-end, although endothelial integrity would recover much sooner than with device
315
implantation. We saw clotting along the cut line in only 1% of our patients a month after
316
surgery, whereas it was observed in 2.2% of patients treated with the Lariat.14 That
317
difference might be because the Lariat technique leaves necrotic LAA tissue, thereby
318
creating longer-lasting inflammatory environment than with our method, in which the
319
LAA is always removed. The Watchman protocol calls for the administration of aspirin
320
after warfarin is discontinued,1 but such a regimen is eliminated with our method. The
321
clinical implications of such use of aspirin are unclear, but it is worth noting Sato et al’s
AC C
EP
TE D
M AN U
310
18
ACCEPTED MANUSCRIPT
322
observation that the use of a low-dose aspirin increases the risk of major bleeding in
323
NVAF.26 Post-procedural pericarditis is not uncommon after the LAA treatment, and we
RI PT
324
observed it in 14% of our patients, as compared with the 8.3% occurrence rate reported
326
with the Lariat technique.14 Pericarditis is not a serious complication, but it can cause
327
prolonged hospitalization or re-admission when the symptoms or pericardial/pleural
328
effusions are severe. Gunda et al. reported that the use of colchicine significantly
329
reduces the incidence and mitigates the symptoms of pericarditis after Lariat ligation.27
330
We adopted a similar protocol and observed improvements in the symptoms and
331
volume of pleural effusions.
M AN U
TE D
332
SC
325
We did not investigate the hemodynamic impact of LAA closure. Atrial natriuretic peptide produced partly by the LAA may decrease with some impact on
334
hemodynamics or diuresis,28 although the Lariat method has been reported not to bring
335
about any changes in secretion.29 With our experience, no patients were re-admitted to
336
hospital for heart failure, and only 1 patient died of chronic heart failure. So far the
337
clinical impact on hemodynamics of the present technique seems negligible, but longer
338
follow-up is necessary to confirm this critical point.
AC C
EP
333
339
19
ACCEPTED MANUSCRIPT
340
STUDY LIMITATIONS
341
The study design was retrospective and not a controlled randomized one, and
RI PT
342
the case volume and follow-up length is insufficient to draw firm conclusions. All patients
344
were operated on by a single surgeon proficient in thoracoscopic surgery, so our clinical
345
outcomes are too limited for us to be able to generalize the procedural efficacy of our
346
method. However, the potential of the procedure was amply demonstrated. Our imaging
347
results, including completeness of the LAA closure, were analyzed by radiologists, but
348
they were not blinded to other clinical data.
M AN U
SC
343
351
352
CONCLUSIONS
EP
350
TE D
349
Our thoracoscopic stapler-and-loop technique swiftly closed the LAA with satisfactory safety and completeness. The 4-year mid-term preventive efficacy (0.25
354
event per 100 patient-years) was acceptable and our brain MRI study provided
355
supporting evidence. Although long-term results are needed before firm conclusions
356
can be drawn, our minimally invasive off-pump technique seems promising as an
357
effective alternative when long-term oral anticoagulation is difficult in patients with
AC C
353
20
ACCEPTED MANUSCRIPT
358
NVAF, particularly in those with an LAA featuring a relatively large size or complex
359
morphology.
ACKNOWLEDGEMENTS
362
363
We thank Randall K. Wolf for his clinical advice.
365
M AN U
364
SC
361
RI PT
360
FUNDING
366
This research did not receive any specific grant from funding agencies in the public,
368
commercial, or not-for-profit sectors.
369
371
REFERENCES
AC C
370
EP
TE D
367
372
1. Reddy VY, Sievert H, Halperin J, et al.; PROTECT AF Steering Committee and
373
Investigators. Percutaneous left atrial appendage closure vs warfarin for atrial
374
fibrillation: a randomized clinical trial. JAMA 2014;312:1988-98.
375
21
ACCEPTED MANUSCRIPT
2. Blackshear JL, Johnson WD, Odell JA, Baker VS, Howard M, Pearce L, Stone C,
377
Packer DL, Schaff HV. Thoracoscopic extracardiac obliteration of the left atrial
378
appendage for stroke risk reduction in atrial fibrillation. J Am Coll Cardiol
379
2003;42:1249-52.
RI PT
376
380
3. Ohtsuka T, Ninomiya M, Nonaka T, Hisagi M, Ota T, Mizutani T. Thoracoscopic
382
stand-alone left atrial appendectomy for thromboembolism prevention in nonvalvular
383
atrial fibrillation. J Am Coll Cardiol 2013;62:103-7.
M AN U
SC
381
384
4. Mokracek A, Kurfirst V, Bulava A, Hanis J, Tesarik R, Pesl L. Thoracoscopic
386
occlusion of the left atrial appendage. Innovations (Phila). 2015;10:179-82.
EP
387
TE D
385
5. Katz ES, Tsiamtsiouris T, Applebaum RM, Schwartzbard A, Tunick PA, Kronzon I.
389
Surgical left atrial appendage ligation is frequently incomplete: a transesophageal
390
echocardiograhic study. J Am Coll Cardiol. 2000;36:468-71.
391
AC C
388
22
ACCEPTED MANUSCRIPT
6. Aryana A, Singh SK, Singh SM, O'Neill PG, Bowers MR, Allen SL, Lewandowski SL,
393
Vierra EC, d'Avila A. Association between incomplete surgical ligation of left atrial
394
appendage and stroke and systemic embolization. Heart Rhythm. 2015;12:1431-7.
RI PT
392
395
7. Lee R, Vassallo P, Kruse J, Malaisrie SC, Rigolin V, Andrei AC, McCarthy P. A
397
randomized, prospective pilot comparison of 3 atrial appendage elimination techniques:
398
Internal ligation, stapled excision, and surgical excision. J Thorac Cardiovasc Surg.
399
2016;152:1075-80.
M AN U
SC
396
400
8. Kanderian AS, Gillinov AM, Pettersson GB, Blackstone E, Klein AL. Success of
402
surgical left atrial appendage closure: assessment by transesophageal
403
echocardiography. J Am Coll Cardiol. 2008;52:924-9.
EP
404
TE D
401
9. Ruttkay T, Scheid M, Götte J, Doll N. Endoscopic resection of a giant left atrial
406
appendage. Innovations (Phila). 2015;10:282-4.
407
AC C
405
23
ACCEPTED MANUSCRIPT
10. Caliskan E, Sahin A, Yilmaz M, et al. Epicardial left atrial appendage AtriClip
409
occlusion reduces the incidence of stroke in patients with atrial fibrillation undergoing
410
cardiac surgery. Europace. 2017 [Epub ahead of print]
RI PT
408
411
11. Osmancik P, Budera P, Zdarska J, Herman D, Petr R, Fojt R, Straka Z. Residual
413
echocardiographic and computed tomography findings after thoracoscopic occlusion of
414
the left atrial appendage using the AtriClip PRO device. Interact Cardiovasc Thorac
415
Surg. 2018 [Epub ahead of print]
416
M AN U
SC
412
12. Boersma LV, Ince H, Kische S, et al. Efficacy and safety of left atrial appendage
418
closure with WATCHMAN in patients with or without contraindication to oral
419
anticoagulation: 1-year follow-up outcome data of the EWOLUTION trial. Heart Rhythm.
420
2017;14:1302-8.
EP
AC C
421
TE D
417
422
13. Tzikas A. Left atrial appendage occlusion with Amplatzer Cardiac Plug and
423
Amplatzer Amulet: a clinical trials update. J Atr Fibrillation. 2017;10:1651.
424
24
ACCEPTED MANUSCRIPT
425
14. Srivastava MC, See VY, Dawood MY, Price MJ. A review of the LARIAT device:
426
insights from the cumulative clinical experience. Springerplus. 2015;4:522.
RI PT
427
15. Reddy VY, Doshi SK, Kar S, Gibson DN, Price MJ, Huber K, Horton RP, Buchbinder
429
M, Neuzil P, Gordon NT, Holmes DR Jr.; PREVAIL and PROTECT AF Investigators.
430
5-Year outcomes after left atrial appendage closure: from the PREVAIL and PROTECT
431
AF trials. J Am Coll Cardiol. 2017;70:2964-75.
M AN U
SC
428
432
16. Bartus K, Morelli RL, Szczepanski W, Kapelak B, Sadowski J, Lee RJ. Anatomic
434
analysis of the left atrial appendage after closure with the LARIAT device. Circ Arrhythm
435
Electrophysiol. 2014;7(4):764–7.
EP
436
TE D
433
17. Burrell LD, Horne BD, Anderson JL, Muhlestein JB, Whisenant BK. Usefulness of
438
left atrial appendage volume as a predictor of embolic stroke in patients with atrial
439
fibrillation. Am J Cardiol 2013;112:1148-52.
440
AC C
437
25
ACCEPTED MANUSCRIPT
18. Hamatani Y, Ogawa H, Takabayashi K, et al. Left atrial enlargement is an
442
independent predictor of stroke and systemic embolism in patients with non-valvular
443
atrial fibrillation. Sci Rep. 2016;6:31042.
RI PT
441
444
19. Di Biase L, Santangeli P, Anselmino M, et al. Does the left atrial appendage
446
morphology correlate with the risk of stroke in patients with atrial fibrillation? Results
447
from a multicenter study. J Am Coll Cardiol 2012;60:531-8.
M AN U
SC
445
448
20. Yamamoto M, Seo Y, Kawamatsu N, Sato K, Sugano A, Machino-Ohtsuka T,
450
Kawamura R, Nakajima H, Igarashi M, Sekiguchi Y, Ishizu T, Aonuma K. Complex left
451
atrial appendage morphology and left atrial appendage thrombus formation in patients
452
with atrial fibrillation. Circ Cardiovasc Imaging. 2014;7:337-43.
EP
TE D
449
453
21. Pillarisetti J, Reddy YM, Gunda S, et al. Endocardial (Watchman) vs epicardial
455
(Lariat) left atrial appendage exclusion devices: understanding the differences in the
456
location and type of leaks and their clinical implications. Heart Rhythm. 2015;12:1501–
457
7.
AC C
454
458
26
ACCEPTED MANUSCRIPT
22. Yarlagadda B, Parikh V, Dar T, Lakkireddy DR. Leaks after left atrial appendage
460
ligation with Lariat device: Incidence, pathophysiology, clinical implications and
461
methods of closure - a case based discussion. Atr Fibrillation. 2017;10:1725.
RI PT
459
462
23. Aryana A, d'Avila A. Incomplete closure of the left atrial appendage: implication and
464
management. Curr Cardiol Rep. 2016;18:82.
M AN U
465
SC
463
24. Yeow WL, Matsumoto T, Kar S. Successful closure of residual leak following
467
LARIAT procedure in a patient with high risk of stroke and hemorrhage. Catheter
468
Cardiovasc Interv. 2014;83:661-3.
TE D
466
469
25. Aznaurov SG, Ball SK, Ellis CR. Thoracoscopic Atriclip closure of left atrial
471
appendage after failed ligation via LARIAT. JACC Cardiovasc Interv. 2015;8:e265-7
AC C
472
EP
470
473
26. Sato H, Ishikawa K, Kitabatake A, et al. Low-dose aspirin for prevention of stroke in
474
low-risk patients with atrial fibrillation: Japan atrial fibrillation stroke trial. Stroke.
475
2006;37:447-51.
476
27
ACCEPTED MANUSCRIPT
27. Gunda S, Reddy M, Nath J, et al. Impact of periprocedural colchicine on
478
postprocedural management in patients undergoing a left atrial appendage ligation
479
using LARIAT. J Cardiovasc Electrophysiol. 2016;27:60-4.
RI PT
477
480
28. Majunke N, Sandri M, Adams V, Daehnert I, Mangner N, Schuler G,
482
Moebius-Winkler S. Atrial and brain natriuretic peptide secretion after percutaneous
483
closure of the left atrial appendage with the Watchman device. J Invasive Cardiol
484
2015;27:448-52.
M AN U
SC
481
485
29. Bartus K, Podolec J, Lee RJ, Kapelak B, Sadowski J, Bartus M, Oles K, Ceranowicz
487
P, Trabka R, Litwinowicz R. Atrial natriuretic peptide and brain natriuretic peptide
488
changes after epicardial percutaneous left atrial appendage suture ligation using
489
LARIAT device. J Physiol Pharmacol 2017;68:117-23.
491 492 493 494
EP
AC C
490
TE D
486
495 496 497 498 499
28
ACCEPTED MANUSCRIPT
Table 1: Patient Characteristics Patient no. (men, women)
201 (118, 83)
Age, mean (SD) [range] years
74 (6) [68-94]
NVAF type, no. (%) 7 (4)
RI PT
Paroxysmal Persistent
16 (8)
Long-standing persistent
178 (88)
NVAF history. no. (%) < 5 years
75 (37)
126 (63)
SC
≥ 5 years
Previous transcutaneous catheter ablation, no. (%)
Mean (SD) [range] Risk criteria, no. (%)
M AN U
CHA2DS2–Vasc score
Congestive heart failure Hypertension Age 65-74 years Age ≥ 75 years
TE D
Diabetes Mellitus
14 (7)
4.1 (1.4) [2-8]
14 (3) 170 (85) 76 (38) 125 (62) 44 (22)
Stroke or transient ischemic attack
77 (38)
Female
91 (45)
Vascular disease Hemodialysis, no. (%)
29 (14) 6 (3)
EP
HAS-BLED score
Mean (SD) [range]
2.9 (1.0) [1-7]
Risk criteria, no. (%)
AC C
500
Hypertension
170 (85)
Renal dysfunction
14 (7)
Liver dysfunction
3 (1)
Stroke or transient ischemic attack
77 (38)
Bleeding
53 (26)
Labile INRs
55 (27)
Age ≥ 65 years
201 (100)
Antiplatelet drugs
9 (4)
Alcohol
2 (1)
29
ACCEPTED MANUSCRIPT Preoperative anticoagulation, no. (%)
186 (93)
Type of anticoagulants, no. (%) Warfarin
38 (20)
Non-vitamin K anticoagulant
148 (80)
Duration, mean (SD) [range] years
2.5 (1.8) [0-10] 53 (26)
Type of anticoagulants, no. (%) Warfarin
12 (23)
Non-vitamin K anticoagulant Site, no. (%)
5 (9)
Gastrointestinal
37 (70)
Others
11 (21)
Median (range) mm
M AN U
Minor side-effects/poor adherence, no. (%)
Left atrial size
41 (77)
SC
Intracranial
Echocardiography
RI PT
Major bleeding events, no. (%)
45 (22)
45 (38-61)
Left ventricular ejection fraction Median (range) mm
56 (38-66)
Apex
TE D
Clots in LAA, by portion, no. (%)
Ostium – mid portion
503 504 505 506 507 508 509
EP
502
NVAF: non-valvular atrial fibrillation.
AC C
501
510 511 512 513 514
30
11 (5) 2 (1)
ACCEPTED MANUSCRIPT
515
Table 2: Surgical outcomes Surgical approach, no. (%) Thoracoscopy
197 (98)
Thoracoscopy + small thoracotomy
4 (2) 0 (0)
Operative time, mean (SD) [range] min
RI PT
Cardiopulmonary bypass, no. (%)
First 100 cases
38 (10) [25-65]
Second 100 cases
28 (7) [16-55]
LAA size (cut-line length in specimen)
48 (35-85)
SC
Median (range) mm Size, no. (%)
≥ 50 mm Hospital death
160 (80)
M AN U
< 50 mm
41 (20) 0 (0)
Major procedure-related events, no. (%) Thromboembolism, no. (%)
0 (0)
Hemorrhage, no. (%)
0 (0)
Phrenic palsy, no. (%)
< 50 ml
TE D
Blood loss, no. (%)
50-200 ml
0 (0)
201 (100) 0 (0) 0 (0)
Reexplorations
0 (0)
Air leakage
6 (3)
EP
Blood transfusions
Aspiration pneumonia
1 (0.5)
AC C
Readmission, no. (%) Bradycardia
1 (0.5)
Pericarditis
1 (0.5)
Post-pericardiotomy syndrome, no. (%)
516
LAA: Left atrial appendage
517 518 519 520 521
31
28 (14)
ACCEPTED MANUSCRIPT
522
FIGURE LEGENDS
523 524
Figure 1: Photographs of 4 thoracoscopic ports. Arrow: Stapler-port.
RI PT
525
Figure 2: Intraoperative 3-dimensional transesophageal echocardiographic images
527
before (left) and after (right) complete closure of left atrial appendage ostium (*) with
528
straight cut-line (arrows).
M AN U
SC
526
529
Figure 3: 3-dimensional left-atrial enhanced computed tomographic images taken
531
before (left) and 1 month (right) after surgery, showing bulky LAA (arrow) was
532
satisfactorily cut and stapled (arrow).
EP
534
AC C
533
TE D
530
32
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S1547-5271(18)30513-7
DOI:
10.1016/j.hrthm.2018.05.026
Reference:
HRTHM 7605
To appear in:
Heart Rhythm
Received Date: 30 March 2018
Please cite this article as: Ohtsuka T, Nonaka T, Hisagi M, Ninomiya M, Masukawa A, Ota T, Thoracoscopic Stapler-and-loop Technique for Left Atrial Appendage Closure in Non-valvular Atrial Fibrillation: Mid-term Outcomes in 201 Patients, Heart Rhythm (2018), doi: 10.1016/j.hrthm.2018.05.026. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
1
Thoracoscopic Stapler-and-loop Technique for Left Atrial Appendage Closure in
2
Non-valvular Atrial Fibrillation: Mid-term Outcomes in 201 Patients
4
RI PT
3
Thoracoscopic stapler-and-loop appendage closure
SC
5
Toshiya Ohtsuka, MD,1 Takahiro Nonaka, MD,1 Motoyuki Hisagi, MD,1
7
Mikio Ninomiya, MD,1 Ai Masukawa, MD,2 Takahiro Ota, MD3
8
COI: None
M AN U
6
9
Departments of Cardiovascular Surgery,1 Radiology2 and Neuro-surgery3
11
Tokyo Metropolitan Tama Medical Center, Japan
TE D
10
EP
12
Corresponding author: Toshiya Ohtsuka
14
Department of Cardiovascular Surgery, Tokyo Metropolitan Tama Medical Center
15
2-8-29 Musashi-dai, Fuchu-shi, Tokyo 183-8524 Japan
16
E-mail: [email protected] TEL: +81-42-323-5111 FAX: +81-42-312-9197
AC C
13
17
18
4996 words
1
ACCEPTED MANUSCRIPT
ABSTRACT
19
20
BACKGROUND: Left atrial appendage (LAA) closure can be an alternative to oral
22
anticoagulation to prevent cardiogenic thromboembolisms in non-valvular atrial
23
fibrillation.
24
OBJECTIVE: We retrospectively evaluated safety, completeness and mid-term
25
prevention of our thoracoscopic stapler-and-loop technique for LAA closure.
26
METHODS: Patients operated on between Oct 2008 and Feb 2017 were reviewed.
27
Endoscopic stapler and ligation-loops were employed. Patients received 1-month of
28
anticoagulation before discontinuation. Hospital death and procedure-related major
29
complications (thromboembolisms, hemorrhagic events, phrenic palsy) were the
30
primary composite endpoint for safety, and cardiogenic thromboembolisms were the
31
endpoint for prevention. Brain magnetic resonance imaging investigated new
32
thromboembolic spots 1 year after surgery.
33
RESULTS: There were 201 patients (118 men, 83 women) with a mean age of 74
34
(range 68-94) years, CHA2DS2–Vasc score 4.1 (SD 1.4), HAS-BLED score 2.9 (SD 1.0).
35
The mean operation time was 28 min. ALL LAAs were removed and intra-operative
36
transesophageal echocardiography confirmed completeness of the closure in each
AC C
EP
TE D
M AN U
SC
RI PT
21
2
ACCEPTED MANUSCRIPT
patient. There were no hospital deaths or major procedure-related complications.
38
Follow-up results for 198 patients (98 %) over a mean period of 48 (range 12-110)
39
months revealed that 2 patients developed cardiogenic thromboembolisms (0.25 event
40
per 100 patient-years). Magnetic resonance imaging of 51 patients with a mean
41
CHA2DS2-Vasc score of 4.7 (SD 1.6) revealed one new small spot in each of 2 patients
42
(3.9%, 3.9 spots per 100 patient-years).
43
CONCLUSIONS: Our thoracoscopic stapler-and-loop technique swiftly, safely and
44
completely closed LAAs in patients with non-valvular atrial fibrillation, and provided
45
acceptable mid-term prevention without anticoagulation.
M AN U
SC
RI PT
37
TE D
46
KEY WORDS: non-valvular atrial fibrillation; cardiogenic thromboembolism; left atrial
48
appendage; stapler; thoracoscopic surgery.
50
51
52
AC C
49
EP
47
53
54
3
ACCEPTED MANUSCRIPT
55
INTRODUCTION
56
Following evidence that a transcutaneous left atrial appendage (LAA) closure is
RI PT
57
as protective as warfarin in non-valvular atrial fibrillation (NVAF),1 new surgical
59
procedures have been invented to close LAAs using minimally invasive off-pump
60
techniques. Some evidence of the efficacy of such techniques exists,2-4 but a better
61
understanding of the potential of minimally invasive surgical options would be of benefit
62
to patients looking for alternatives to long-term oral anticoagulation.
M AN U
63
SC
58
We reviewed our 9-year clinical experience of a port-access thoracoscopic stapler-and-loop technique for LAA closure in patients with NVAF to evaluate its safety,
65
completeness and mid-term prevention of cardiogenic thromboembolisms.
66
68
69
METHODS
AC C
67
EP
TE D
64
Our institutional review board approved the clinical protocol for the procedure in
70
2008, and written informed consent to surgery and procedure-related examinations
71
were obtained from each patient.
72
4
ACCEPTED MANUSCRIPT
73
Patient Selection The present technique was offered to patients at risk of cardiogenic
75
thromboembolisms (CHA2DS2-Vasc score ≥ 1) when long-term anticoagulation was
76
deemed difficult (HAS-BLED score ≥ 1) for them, or they had experienced cardiogenic
77
thromboembolisms despite receiving a moderate to high doses of anticoagulants. The
78
procedure was also considered for patients with poor adherence to anticoagulation
79
regimens because, for example, of dementia. Preoperative screening was used to
80
demonstrate that NVAF was the only risk factor for thromboembolisms, and to show that
81
it would be difficult to sustain sinus rhythm with conservative or invasive anti-arrhythmic
82
treatments.
SC
M AN U
TE D
83
RI PT
74
Excluded from the operation were patients deemed unlikely to be able to tolerate 1-month of anticoagulation, those who had a significantly low left-ventricular
85
ejection fraction (< 25 %), those with an inaccessible left thoracic cavity due to
86
pleural/pericardial adhesions, and those unlikely to be able to tolerate a left-lung
87
collapse of longer than 30 min, e.g. patients with terminal pulmonary emphysema. If the
88
LAA contained a clot except in its apical portion, the operation was postponed. Neither
89
the size nor morphology of the LAA limited eligibility for the procedure.
AC C
EP
84
5
ACCEPTED MANUSCRIPT
90
Neurological screening test of the carotid artery was performed; if any lesions were discovered, neurologists followed up and conducted conservative or surgical
92
treatments if necessary.
93
95
Surgical Method (Supplementary Video) and Clinical Pathway
General anesthesia was established through a double-lumen endotracheal
SC
94
RI PT
91
tube. Transesophageal echocardiography (TEE) checked the LAA for clot formation.
97
Four ports were placed in the left lateral thoracic region (Figure 1). No heparin was used
98
during or after surgery. We employed an automatic, bendable cut-and-staple device
99
(ECHELON FLEX™ Powered ENDOPATH® Stapler 60, ETHICON, USA) which can cut
100
and close 6-cm-long tissue with a single cartridge. If any small protrusion was observed
101
at either end of the staple line, an endoscopic ligation tool (ENDOLOOP® Ligature
102
EZ10G, ETHICON, USA) was used to obliterate it. Intraoperative TEE was used for
103
quality control: it guided LAA closure, and recorded the process and results (Figure 2).
105
TE D
EP
AC C
104
M AN U
96
In accordance with our clinical pathway, patients were discharged from hospital
at postoperative day 3.
106
107
6
ACCEPTED MANUSCRIPT
108
109
Anticoagulation Management Anticoagulation was discontinued in accordance with our institutional protocol as follows. Any anticoagulant treatment was suspended 1-2 days before surgery.
111
Anticoagulation (with the same anticoagulant as used preoperatively) was resumed
112
immediately after surgery and continued for 1 month. Enhanced computed tomography
113
was carried out one-month-postoperatively to examine the left atrium, and
114
anticoagulation was thereafter discontinued if the examination proved that the LAA had
115
been successfully removed without a stump (≥ 10 mm in depth) and that there was no
116
clotting along the cut-line. No antiplatelet regimens were used unless patients were
117
already taking them for arteriosclerotic diseases.
SC
M AN U
TE D
119
Endpoints for Procedural Safety
EP
118
RI PT
110
We reviewed hospital charts to identify hospital deaths and major
121
procedure-related complications as the primary composite endpoints for procedural
122
safety, and other complications as secondary ones. Three serious events were defined
123
as major: thromboembolisms, hemorrhagic events and permanent phrenic palsy caused
124
by phrenic nerve injury during pericardiotomy.
AC C
120
125
7
ACCEPTED MANUSCRIPT
126
127
Endpoints for Prevention against Cardiogenic Thromboembolism Neurological investigations were conducted 1, 3, 6 and 12 months after surgery, and yearly thereafter either in the outpatient clinic or by phone or e-mail. Whenever
129
patients had clinical events suggesting ischemia of the brain (disturbance of
130
consciousness, or motor or sensory disorders) or other organs, they were examined
131
and treated. Stroke, transient ischemic attack (TIA) and other ischemic symptoms were
132
judged to be cardiogenic thromboembolisms unless other clear etiologies existed. Using
133
the follow-up data, we identified cardiogenic thromboembolisms and estimated the
134
event rates per 100 patient-years.
137
Brain Magnetic Resonance Imaging Study
At 1-year follow-up, magnetic resonance imaging (MRI) was used to identify
EP
136
TE D
135
M AN U
SC
RI PT
128
new cerebral spots. For this purpose, we enrolled those patients who had experienced
139
stroke or TIA preoperatively and had undergone the diagnostic brain MRI. The
140
preoperative MRI was used as a control and new cerebral spots were counted in the
141
1-year-follow-up images.
AC C
138
142
143
8
ACCEPTED MANUSCRIPT
144
RESULTS
145
147
Patients Characteristics (Table 1)
RI PT
146
Between October 2008 and February 2017, 201 (118 men, 83 women) patients were operated on (3 target patients were excluded). The mean age was 74 years old
149
(range 68-94). NVAF was paroxysmal in 7 patients (4%), persistent in 16 (8%), and
150
long-standing persistent in 178 (88%). 126 patients (63%) had had NVAF for 5 years or
151
longer. The mean CHA2DS2-Vasc and HAS-BLED scores were 4.1 (SD 1.4) and 2.9
152
(SD 1.0), respectively. Nine patients (4%) had been on antiplatelet drugs for
153
arteriosclerotic diseases. Anticoagulants had been used for a mean period of 2.5 years
154
(SD 1.8) [range 0-10] to treat 186 (93%) of the patients (warfarin for 38 of them, and
155
non-vitamin K anticoagulants for 148). Of these 186 patients, 53 (26%) had experienced
156
anticoagulation-associated major bleeding: 5 intracranial, 37 gastrointestinal and 11
157
other hemorrhages. The median left atrial size was 45 mm (range 38-61) on
158
echocardiography. Clot formation had been suspected exclusively at the apical portion
159
of the LAA in 11 patients (5%), although all of them had been continuously
160
anticoagulated. Clots had been positively identified near the ostium or at the mid-portion
161
of the LAA in 2 patients (1%), both of whom had experienced cardiogenic
AC C
EP
TE D
M AN U
SC
148
9
ACCEPTED MANUSCRIPT
thromboembolisms shortly before the examinations; they were therefore treated with
163
warfarin to maintain relatively high INRs for a while (3 months at the longest) until
164
reexamination confirmed the absence of clots; they then underwent surgery.
RI PT
162
165
167
Surgical Outcomes (Table 2)
A single surgeon operated on all patients at a single institute. The approach
SC
166
was port-access thoracoscopy alone in 197 (98%) of the operations, and video-assisted
169
small thoracotomy in the others, because the patients had pericardial or pleural scars.
170
None of the patients underwent cardiopulmonary bypass. The mean operation time for
171
the first half of the patients was 38 min (SD 10) and the second half 28 min (SD 7). The
172
staple-and-loop technique was used to close all LAAs. The completeness of each LAA
173
closure was immediately confirmed with intraoperative TEE. The median cut-length of
174
the LAA was 48 mm (range 35-85 mm).
TE D
EP
There were no hospital deaths and no major procedure-related complications.
AC C
175
M AN U
168
176
Blood loss was less than 50 ml in all operations. Six patients (3.0%) had persistent but
177
minor pulmonary air-leakage, and 1 (0.5%) with dysphasia due to a preoperative stroke
178
suffered aspiration pneumonia. There were 2 re-admissions (1.0%): one patient
179
gradually developed bradycardia, which was treated with a pace-maker implantation,
10
ACCEPTED MANUSCRIPT
and the other developed serious pericarditis, which was treated with steroid
181
administration. 28 patients (14%) suffered post-pericardiotomy syndrome with typical
182
symptoms (chest pain, dry cough, pleural effusions); all symptoms occurred less than 1
183
month after surgery and were abated conservatively with non-steroid anti-inflammatory
184
medication.
SC
RI PT
180
185
187
Follow-up Outcomes
M AN U
186
Computed tomography was carried out 1 month postoperatively in all but 4 (2.0%) of the patients. Three patients (1.5 %) in our early cases revealed a stump
189
(greater than 5 mm but less than 1 cm in depth), and 4 patients (2%) who were operated
190
on before we adopted the endo-loop technique showed a narrow protrusion (shorter
191
than 5 mm) at the caudal end of the staple line. In all of the other patients, the LAA was
192
flat or formed a gentle convexity (Figure 3). A left-atrial clot along the cut line was
193
observed in 2 patients (1.0%); they took anticoagulants for another 2 months, and
194
further examinations revealed 6 months postoperatively that the clots had disappeared.
195
Thus, except for these 2 cases, all of the patients discontinued anticoagulation no later
196
than 1 month after surgery.
AC C
EP
TE D
188
11
ACCEPTED MANUSCRIPT
197
Three patients (1.5%) with CHA2DS2-Vasc scores of 3, 3 and 4, respectively, were lost to follow-up, so outcomes were obtained for 198 patients. The mean and
199
median follow-up periods were 48 (SD 25) [range 12-113] and 46 months. Aggregate
200
patient-years reached 801. In the entire follow-up periods, 5 neurological events
201
occurred, but there were no extra-cerebral thromboembolisms. Three events were
202
proven to be not due to cardiogenic thromboembolisms: 1 patient developed TIA as a
203
stenotic lesion at the carotid artery gradually worsened and therefore underwent an
204
endarterectomy; another patient developed ischemic symptoms due to a known left
205
vertebral artery stenosis and received a bypass operation; and the other patient
206
suffered hypertensive cerebellar bleeding which was treated conservatively. The other 2
207
events were judged to be cardiogenic thromboembolisms (1.0%, 0.25 event per 100
208
patient-years), although satisfactory closure of the LAA had been exhibited in
209
1-month-postoperative enhanced computed tomography. One was TIA (rated modified
210
Rankin score 1), which occurred 2 years after surgery in a 72-year-old woman with
211
long-standing persistent NVAF, a CHA2DS2-Vasc score of 3, and a left atrial size of 45
212
mm. The other event was stroke (rated modified Rankin score 1), which struck 5 months
213
after surgery in a 66-year-old man with long-standing persistent NVAF, a
214
CHA2DS2-Vasc score of 3, and a left atrial size of 50 mm. This was only patient who
AC C
EP
TE D
M AN U
SC
RI PT
198
12
ACCEPTED MANUSCRIPT
215
remained on warfarin maintaining the INR between 2.0 and 2.5 till the end of follow-up.
216
These 5 patients remained free from neurological symptoms thereafter. Four of the patients died (2.0%, 0.53 event per 100 patient-years), 2 of
RI PT
217
malignant or non-cardiac chronic diseases, 1 of chronic heart failure, and 1 of acute
219
coronary syndrome due to a plaque rupture in the left main coronary artery. They were
220
free from cardiogenic thromboembolisms in their follow-up periods.
SC
218
222
Brain-MRI Study Results
M AN U
221
51 patients underwent preoperative brain MRI. The mean age was 75 years
224
(SD 6), with a range of 70-84, and the mean CHA2DS2-Vasc score was 4.7 (SD 1.6),
225
with a range of 2-8. Brain MRI carried out 1 year after surgery revealed 1 new spot in
226
each of only 2 patients (1.0 %, 3.8 spots per 100 patient-years). Each spot was at the
227
cerebral cortex and smaller than 5 mm in diameter.
229
230
EP
AC C
228
TE D
223
231
232
13
ACCEPTED MANUSCRIPT
233
DISCUSSION
234
In 2013, we published a feasibility study of the present technique elsewhere.3
RI PT
235
236
The present report concerns the procedural efficacies: safety, completeness and
237
mid-term prevention of cardiogenic thromboembolisms..
Three kinds of surgical technique for epicardial LAA closure have been
SC
238
previously reported in the literature: suture ligation, stapling, and clipping of the LAA.
240
Suture ligation is the simplest and cheapest method, but the reported results
M AN U
239
were unfavorable in terms of completeness and stroke prevention. Katz et al. found
242
36% of the LAAs were incompletely ligated in concomitant heart surgeries.5 Aryana et al.
243
reported a significantly increased risk (19% per 100 patient-years of follow-up) of stroke
244
or systemic embolization when the neck of the patent LAA was 5 mm or less in
245
diameter.6 No reports of thoracoscopic suture ligation of the LAA have been published
246
since Blackshear et al. reported on 15 patients in 2003,2 so surgeons would probably be
247
well advised to avoid this technique (either via open or thoracoscopic approach) except
248
for the LAAs with a very small, round ostium.
249
250
AC C
EP
TE D
241
Although the endo-stapler is a commonly used and, therefore, reasonably priced device, some surgeons have questioned its performance in LAA closure. Lee et
14
ACCEPTED MANUSCRIPT
al. compared 3 different techniques and found that stapling left unacceptable residual
252
stumps in 60% of their subjects.7 Kanderian et al. reported no success with the device.8
253
One possible reason for such failure may be that few heart surgeons are familiar with
254
endo-staplers, and another likely reason is that LAA closure was conducted via
255
sternotomy: the LAA is deeply located in the pericardial cavity via sternotomy, so it is
256
difficult to direct the device properly along the ostium of the LAA. Besides, viewed from
257
above, its cranial part is often hidden behind the left pulmonary artery, so there is a
258
residual stump there. On the other hand, thoracoscopy allows visualization of all parts of
259
the LAA, and a stapler can be properly introduced to the LAA base through a port in the
260
lateral thoracic region. Ruttkay et al. completely closed a giant LAA with a similar
261
approach to ours.9
SC
M AN U
TE D
Clipping is the most expensive method, but clipping devices are remarkably
EP
262
RI PT
251
attractive because they permit atraumatic and arguably more surgeon-friendly
264
techniques than stapling of the LAA. Caliskan et al. reported their clinical experience in
265
open heart surgery in which all LAAs were clipped with no sign of significant stumps and
266
satisfactory mid-term results in terms of stroke prevention (0.5 per 100 patient-years),10
267
a similar result to ours. Osmancik et al. reported a very high success rate (97.5%) in 40
268
cases of thoracoscopic clip ligation, but they found stumps deeper than 10 mm in 45%
AC C
263
15
ACCEPTED MANUSCRIPT of the patients, and a volume of 3.6 ± 1.9 ml.11 The clinical implications of such residual
270
stumps are unclear, but the quality of closure seems to be compromised by use of the
271
device off pump. Once the clip has been deployed on the LAA, its position can never be
272
adjusted and, unlike with our endo-loop technique, no touch-up measures are available
273
to obliterate below-clip stumps. This kind of problem needs to be overcome by further
274
material improvements.
SC
In the field of LAA management, transcutaneous catheter techniques have also
M AN U
275
RI PT
269
evolved although they are before approval in Japan: device implantation into the LAA
277
with occluders, e.g. the Watchman, and extracardiac ligation with the Lariat device.12-14
278
These techniques are no doubt less invasive (or less incisional) than any surgical
279
methods, but have controversial issues to be discussed in comparison with our method.
280
First, procedural safety is a pivotal point. While Watchman implantation seems
EP
TE D
276
to have become less risky as far as the results of the PREVAIL versus PROTECT AF
282
study are concerned,15 the Lariat technique was shown by Srivasta et al. to have yielded
283
suboptimal outcomes, causing cardiac tamponade in 7.5% of patients, LAA laceration in
284
1.8%, and death in 0.3%.14 Thoracoscopy clearly shows that LAA tissue is thick and
285
strong at the basal part, but very thin and breakable at the other parts; we therefore
AC C
281
16
ACCEPTED MANUSCRIPT
never pinch and pull up the LAA with our technique. Attention needs to be paid to this
287
anatomical characteristic to further improve the safety of the transcutaneous methods.
288
The second issue is the procedural limitations dictated by the LAA’s size and
RI PT
286
morphology. Although our method is not limited according to LAA anatomy, the
290
transcutaneous techniques certainly are. According to the instruction for use, the
291
Watchman cannot be implanted if the maximum ostial width exceeds 31 mm. Bartus et
292
al. recommended the Lariat technique not be used when the LAA is wider than 40 mm,
293
when the LAA is superiorly oriented with its apex behind the pulmonary trunk, and when
294
the lobes in a multi-lobed LAA are oriented in different planes exceeding 40 mm.16 No
295
reports on either technique have clarified what percentage of LAAs are actually
296
excluded because of their size or morphology, although it was clear from the Protect AF
297
study that 1945/4998 cases (39%) were excluded for clinical or echocardiographic
298
reasons.1 Since previous reports show that LAAs with a larger size or more complex
299
shape pose a greater thromboembolic risk to patients,17-20 we believe that our technique
300
should treat LAAs with such risky anatomies, rather than transcutaneous methods.
M AN U
TE D
EP
AC C
301
SC
289
Residual gaps are also an issue. Although we remove the LAA in every patient,
302
after 1 year of follow-up, Pillarisetti et al. reported that the Watchman left a gap in 21%
303
of subject patients, and the Lariat device in 14%.21 Yarlagadda et al. stressed that
17
ACCEPTED MANUSCRIPT
ligation with the Lariat created fewer leaks than surgical ligation, and that the recurrent
305
gaps had no link with cardiogenic thromboembolic events after either technique.22
306
Nevertheless, as Aryana warned,23 the patent LAA is able to generate thrombo-emboli.
307
Plugs or clips might be able to control central leaks after ligation with the Lariat,24,25 but it
308
would be extremely difficult to stop edge leaks after device implantation.
After Watchman implantation, anticoagulation management continues until the
SC
309
RI PT
304
endothelial coverage of the device is completed, although progression varies from
311
patient to patient.1 Our protocol might be simpler: anticoagulants are administered for 1
312
month postoperatively. We established this protocol because procedure-associated
313
inflammatory reactions in the acute post-operative period might cause clotting at the
314
cut-end, although endothelial integrity would recover much sooner than with device
315
implantation. We saw clotting along the cut line in only 1% of our patients a month after
316
surgery, whereas it was observed in 2.2% of patients treated with the Lariat.14 That
317
difference might be because the Lariat technique leaves necrotic LAA tissue, thereby
318
creating longer-lasting inflammatory environment than with our method, in which the
319
LAA is always removed. The Watchman protocol calls for the administration of aspirin
320
after warfarin is discontinued,1 but such a regimen is eliminated with our method. The
321
clinical implications of such use of aspirin are unclear, but it is worth noting Sato et al’s
AC C
EP
TE D
M AN U
310
18
ACCEPTED MANUSCRIPT
322
observation that the use of a low-dose aspirin increases the risk of major bleeding in
323
NVAF.26 Post-procedural pericarditis is not uncommon after the LAA treatment, and we
RI PT
324
observed it in 14% of our patients, as compared with the 8.3% occurrence rate reported
326
with the Lariat technique.14 Pericarditis is not a serious complication, but it can cause
327
prolonged hospitalization or re-admission when the symptoms or pericardial/pleural
328
effusions are severe. Gunda et al. reported that the use of colchicine significantly
329
reduces the incidence and mitigates the symptoms of pericarditis after Lariat ligation.27
330
We adopted a similar protocol and observed improvements in the symptoms and
331
volume of pleural effusions.
M AN U
TE D
332
SC
325
We did not investigate the hemodynamic impact of LAA closure. Atrial natriuretic peptide produced partly by the LAA may decrease with some impact on
334
hemodynamics or diuresis,28 although the Lariat method has been reported not to bring
335
about any changes in secretion.29 With our experience, no patients were re-admitted to
336
hospital for heart failure, and only 1 patient died of chronic heart failure. So far the
337
clinical impact on hemodynamics of the present technique seems negligible, but longer
338
follow-up is necessary to confirm this critical point.
AC C
EP
333
339
19
ACCEPTED MANUSCRIPT
340
STUDY LIMITATIONS
341
The study design was retrospective and not a controlled randomized one, and
RI PT
342
the case volume and follow-up length is insufficient to draw firm conclusions. All patients
344
were operated on by a single surgeon proficient in thoracoscopic surgery, so our clinical
345
outcomes are too limited for us to be able to generalize the procedural efficacy of our
346
method. However, the potential of the procedure was amply demonstrated. Our imaging
347
results, including completeness of the LAA closure, were analyzed by radiologists, but
348
they were not blinded to other clinical data.
M AN U
SC
343
351
352
CONCLUSIONS
EP
350
TE D
349
Our thoracoscopic stapler-and-loop technique swiftly closed the LAA with satisfactory safety and completeness. The 4-year mid-term preventive efficacy (0.25
354
event per 100 patient-years) was acceptable and our brain MRI study provided
355
supporting evidence. Although long-term results are needed before firm conclusions
356
can be drawn, our minimally invasive off-pump technique seems promising as an
357
effective alternative when long-term oral anticoagulation is difficult in patients with
AC C
353
20
ACCEPTED MANUSCRIPT
358
NVAF, particularly in those with an LAA featuring a relatively large size or complex
359
morphology.
ACKNOWLEDGEMENTS
362
363
We thank Randall K. Wolf for his clinical advice.
365
M AN U
364
SC
361
RI PT
360
FUNDING
366
This research did not receive any specific grant from funding agencies in the public,
368
commercial, or not-for-profit sectors.
369
371
REFERENCES
AC C
370
EP
TE D
367
372
1. Reddy VY, Sievert H, Halperin J, et al.; PROTECT AF Steering Committee and
373
Investigators. Percutaneous left atrial appendage closure vs warfarin for atrial
374
fibrillation: a randomized clinical trial. JAMA 2014;312:1988-98.
375
21
ACCEPTED MANUSCRIPT
2. Blackshear JL, Johnson WD, Odell JA, Baker VS, Howard M, Pearce L, Stone C,
377
Packer DL, Schaff HV. Thoracoscopic extracardiac obliteration of the left atrial
378
appendage for stroke risk reduction in atrial fibrillation. J Am Coll Cardiol
379
2003;42:1249-52.
RI PT
376
380
3. Ohtsuka T, Ninomiya M, Nonaka T, Hisagi M, Ota T, Mizutani T. Thoracoscopic
382
stand-alone left atrial appendectomy for thromboembolism prevention in nonvalvular
383
atrial fibrillation. J Am Coll Cardiol 2013;62:103-7.
M AN U
SC
381
384
4. Mokracek A, Kurfirst V, Bulava A, Hanis J, Tesarik R, Pesl L. Thoracoscopic
386
occlusion of the left atrial appendage. Innovations (Phila). 2015;10:179-82.
EP
387
TE D
385
5. Katz ES, Tsiamtsiouris T, Applebaum RM, Schwartzbard A, Tunick PA, Kronzon I.
389
Surgical left atrial appendage ligation is frequently incomplete: a transesophageal
390
echocardiograhic study. J Am Coll Cardiol. 2000;36:468-71.
391
AC C
388
22
ACCEPTED MANUSCRIPT
6. Aryana A, Singh SK, Singh SM, O'Neill PG, Bowers MR, Allen SL, Lewandowski SL,
393
Vierra EC, d'Avila A. Association between incomplete surgical ligation of left atrial
394
appendage and stroke and systemic embolization. Heart Rhythm. 2015;12:1431-7.
RI PT
392
395
7. Lee R, Vassallo P, Kruse J, Malaisrie SC, Rigolin V, Andrei AC, McCarthy P. A
397
randomized, prospective pilot comparison of 3 atrial appendage elimination techniques:
398
Internal ligation, stapled excision, and surgical excision. J Thorac Cardiovasc Surg.
399
2016;152:1075-80.
M AN U
SC
396
400
8. Kanderian AS, Gillinov AM, Pettersson GB, Blackstone E, Klein AL. Success of
402
surgical left atrial appendage closure: assessment by transesophageal
403
echocardiography. J Am Coll Cardiol. 2008;52:924-9.
EP
404
TE D
401
9. Ruttkay T, Scheid M, Götte J, Doll N. Endoscopic resection of a giant left atrial
406
appendage. Innovations (Phila). 2015;10:282-4.
407
AC C
405
23
ACCEPTED MANUSCRIPT
10. Caliskan E, Sahin A, Yilmaz M, et al. Epicardial left atrial appendage AtriClip
409
occlusion reduces the incidence of stroke in patients with atrial fibrillation undergoing
410
cardiac surgery. Europace. 2017 [Epub ahead of print]
RI PT
408
411
11. Osmancik P, Budera P, Zdarska J, Herman D, Petr R, Fojt R, Straka Z. Residual
413
echocardiographic and computed tomography findings after thoracoscopic occlusion of
414
the left atrial appendage using the AtriClip PRO device. Interact Cardiovasc Thorac
415
Surg. 2018 [Epub ahead of print]
416
M AN U
SC
412
12. Boersma LV, Ince H, Kische S, et al. Efficacy and safety of left atrial appendage
418
closure with WATCHMAN in patients with or without contraindication to oral
419
anticoagulation: 1-year follow-up outcome data of the EWOLUTION trial. Heart Rhythm.
420
2017;14:1302-8.
EP
AC C
421
TE D
417
422
13. Tzikas A. Left atrial appendage occlusion with Amplatzer Cardiac Plug and
423
Amplatzer Amulet: a clinical trials update. J Atr Fibrillation. 2017;10:1651.
424
24
ACCEPTED MANUSCRIPT
425
14. Srivastava MC, See VY, Dawood MY, Price MJ. A review of the LARIAT device:
426
insights from the cumulative clinical experience. Springerplus. 2015;4:522.
RI PT
427
15. Reddy VY, Doshi SK, Kar S, Gibson DN, Price MJ, Huber K, Horton RP, Buchbinder
429
M, Neuzil P, Gordon NT, Holmes DR Jr.; PREVAIL and PROTECT AF Investigators.
430
5-Year outcomes after left atrial appendage closure: from the PREVAIL and PROTECT
431
AF trials. J Am Coll Cardiol. 2017;70:2964-75.
M AN U
SC
428
432
16. Bartus K, Morelli RL, Szczepanski W, Kapelak B, Sadowski J, Lee RJ. Anatomic
434
analysis of the left atrial appendage after closure with the LARIAT device. Circ Arrhythm
435
Electrophysiol. 2014;7(4):764–7.
EP
436
TE D
433
17. Burrell LD, Horne BD, Anderson JL, Muhlestein JB, Whisenant BK. Usefulness of
438
left atrial appendage volume as a predictor of embolic stroke in patients with atrial
439
fibrillation. Am J Cardiol 2013;112:1148-52.
440
AC C
437
25
ACCEPTED MANUSCRIPT
18. Hamatani Y, Ogawa H, Takabayashi K, et al. Left atrial enlargement is an
442
independent predictor of stroke and systemic embolism in patients with non-valvular
443
atrial fibrillation. Sci Rep. 2016;6:31042.
RI PT
441
444
19. Di Biase L, Santangeli P, Anselmino M, et al. Does the left atrial appendage
446
morphology correlate with the risk of stroke in patients with atrial fibrillation? Results
447
from a multicenter study. J Am Coll Cardiol 2012;60:531-8.
M AN U
SC
445
448
20. Yamamoto M, Seo Y, Kawamatsu N, Sato K, Sugano A, Machino-Ohtsuka T,
450
Kawamura R, Nakajima H, Igarashi M, Sekiguchi Y, Ishizu T, Aonuma K. Complex left
451
atrial appendage morphology and left atrial appendage thrombus formation in patients
452
with atrial fibrillation. Circ Cardiovasc Imaging. 2014;7:337-43.
EP
TE D
449
453
21. Pillarisetti J, Reddy YM, Gunda S, et al. Endocardial (Watchman) vs epicardial
455
(Lariat) left atrial appendage exclusion devices: understanding the differences in the
456
location and type of leaks and their clinical implications. Heart Rhythm. 2015;12:1501–
457
7.
AC C
454
458
26
ACCEPTED MANUSCRIPT
22. Yarlagadda B, Parikh V, Dar T, Lakkireddy DR. Leaks after left atrial appendage
460
ligation with Lariat device: Incidence, pathophysiology, clinical implications and
461
methods of closure - a case based discussion. Atr Fibrillation. 2017;10:1725.
RI PT
459
462
23. Aryana A, d'Avila A. Incomplete closure of the left atrial appendage: implication and
464
management. Curr Cardiol Rep. 2016;18:82.
M AN U
465
SC
463
24. Yeow WL, Matsumoto T, Kar S. Successful closure of residual leak following
467
LARIAT procedure in a patient with high risk of stroke and hemorrhage. Catheter
468
Cardiovasc Interv. 2014;83:661-3.
TE D
466
469
25. Aznaurov SG, Ball SK, Ellis CR. Thoracoscopic Atriclip closure of left atrial
471
appendage after failed ligation via LARIAT. JACC Cardiovasc Interv. 2015;8:e265-7
AC C
472
EP
470
473
26. Sato H, Ishikawa K, Kitabatake A, et al. Low-dose aspirin for prevention of stroke in
474
low-risk patients with atrial fibrillation: Japan atrial fibrillation stroke trial. Stroke.
475
2006;37:447-51.
476
27
ACCEPTED MANUSCRIPT
27. Gunda S, Reddy M, Nath J, et al. Impact of periprocedural colchicine on
478
postprocedural management in patients undergoing a left atrial appendage ligation
479
using LARIAT. J Cardiovasc Electrophysiol. 2016;27:60-4.
RI PT
477
480
28. Majunke N, Sandri M, Adams V, Daehnert I, Mangner N, Schuler G,
482
Moebius-Winkler S. Atrial and brain natriuretic peptide secretion after percutaneous
483
closure of the left atrial appendage with the Watchman device. J Invasive Cardiol
484
2015;27:448-52.
M AN U
SC
481
485
29. Bartus K, Podolec J, Lee RJ, Kapelak B, Sadowski J, Bartus M, Oles K, Ceranowicz
487
P, Trabka R, Litwinowicz R. Atrial natriuretic peptide and brain natriuretic peptide
488
changes after epicardial percutaneous left atrial appendage suture ligation using
489
LARIAT device. J Physiol Pharmacol 2017;68:117-23.
491 492 493 494
EP
AC C
490
TE D
486
495 496 497 498 499
28
ACCEPTED MANUSCRIPT
Table 1: Patient Characteristics Patient no. (men, women)
201 (118, 83)
Age, mean (SD) [range] years
74 (6) [68-94]
NVAF type, no. (%) 7 (4)
RI PT
Paroxysmal Persistent
16 (8)
Long-standing persistent
178 (88)
NVAF history. no. (%) < 5 years
75 (37)
126 (63)
SC
≥ 5 years
Previous transcutaneous catheter ablation, no. (%)
Mean (SD) [range] Risk criteria, no. (%)
M AN U
CHA2DS2–Vasc score
Congestive heart failure Hypertension Age 65-74 years Age ≥ 75 years
TE D
Diabetes Mellitus
14 (7)
4.1 (1.4) [2-8]
14 (3) 170 (85) 76 (38) 125 (62) 44 (22)
Stroke or transient ischemic attack
77 (38)
Female
91 (45)
Vascular disease Hemodialysis, no. (%)
29 (14) 6 (3)
EP
HAS-BLED score
Mean (SD) [range]
2.9 (1.0) [1-7]
Risk criteria, no. (%)
AC C
500
Hypertension
170 (85)
Renal dysfunction
14 (7)
Liver dysfunction
3 (1)
Stroke or transient ischemic attack
77 (38)
Bleeding
53 (26)
Labile INRs
55 (27)
Age ≥ 65 years
201 (100)
Antiplatelet drugs
9 (4)
Alcohol
2 (1)
29
ACCEPTED MANUSCRIPT Preoperative anticoagulation, no. (%)
186 (93)
Type of anticoagulants, no. (%) Warfarin
38 (20)
Non-vitamin K anticoagulant
148 (80)
Duration, mean (SD) [range] years
2.5 (1.8) [0-10] 53 (26)
Type of anticoagulants, no. (%) Warfarin
12 (23)
Non-vitamin K anticoagulant Site, no. (%)
5 (9)
Gastrointestinal
37 (70)
Others
11 (21)
Median (range) mm
M AN U
Minor side-effects/poor adherence, no. (%)
Left atrial size
41 (77)
SC
Intracranial
Echocardiography
RI PT
Major bleeding events, no. (%)
45 (22)
45 (38-61)
Left ventricular ejection fraction Median (range) mm
56 (38-66)
Apex
TE D
Clots in LAA, by portion, no. (%)
Ostium – mid portion
503 504 505 506 507 508 509
EP
502
NVAF: non-valvular atrial fibrillation.
AC C
501
510 511 512 513 514
30
11 (5) 2 (1)
ACCEPTED MANUSCRIPT
515
Table 2: Surgical outcomes Surgical approach, no. (%) Thoracoscopy
197 (98)
Thoracoscopy + small thoracotomy
4 (2) 0 (0)
Operative time, mean (SD) [range] min
RI PT
Cardiopulmonary bypass, no. (%)
First 100 cases
38 (10) [25-65]
Second 100 cases
28 (7) [16-55]
LAA size (cut-line length in specimen)
48 (35-85)
SC
Median (range) mm Size, no. (%)
≥ 50 mm Hospital death
160 (80)
M AN U
< 50 mm
41 (20) 0 (0)
Major procedure-related events, no. (%) Thromboembolism, no. (%)
0 (0)
Hemorrhage, no. (%)
0 (0)
Phrenic palsy, no. (%)
< 50 ml
TE D
Blood loss, no. (%)
50-200 ml
0 (0)
201 (100) 0 (0) 0 (0)
Reexplorations
0 (0)
Air leakage
6 (3)
EP
Blood transfusions
Aspiration pneumonia
1 (0.5)
AC C
Readmission, no. (%) Bradycardia
1 (0.5)
Pericarditis
1 (0.5)
Post-pericardiotomy syndrome, no. (%)
516
LAA: Left atrial appendage
517 518 519 520 521
31
28 (14)
ACCEPTED MANUSCRIPT
522
FIGURE LEGENDS
523 524
Figure 1: Photographs of 4 thoracoscopic ports. Arrow: Stapler-port.
RI PT
525
Figure 2: Intraoperative 3-dimensional transesophageal echocardiographic images
527
before (left) and after (right) complete closure of left atrial appendage ostium (*) with
528
straight cut-line (arrows).
M AN U
SC
526
529
Figure 3: 3-dimensional left-atrial enhanced computed tomographic images taken
531
before (left) and 1 month (right) after surgery, showing bulky LAA (arrow) was
532
satisfactorily cut and stapled (arrow).
EP
534
AC C
533
TE D
530
32
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT