If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
St. Antonius Hospital, Cardiology Department, Koekoekslaan 1, 3455CM, Nieuwegein, the NetherlandsAmsterdam UMC, Cardiology Department, Meibergdreef 9, 1105 AZ, Amsterdam, Netherlands
St. Antonius Hospital, Cardiology Department, Koekoekslaan 1, 3455CM, Nieuwegein, the NetherlandsAmsterdam UMC, Cardiology Department, Meibergdreef 9, 1105 AZ, Amsterdam, Netherlands
St. Antonius Hospital, Cardiology Department, Koekoekslaan 1, 3455CM, Nieuwegein, the NetherlandsAmsterdam UMC, Cardiology Department, Meibergdreef 9, 1105 AZ, Amsterdam, Netherlands
St. Antonius Hospital, Cardiology Department, Koekoekslaan 1, 3455CM, Nieuwegein, the NetherlandsAmsterdam UMC, Cardiology Department, Meibergdreef 9, 1105 AZ, Amsterdam, Netherlands
St. Antonius Hospital, Cardiology Department, Koekoekslaan 1, 3455CM, Nieuwegein, the NetherlandsAmsterdam UMC, Cardiology Department, Meibergdreef 9, 1105 AZ, Amsterdam, Netherlands
Percutaneous left atrial appendage occlusion (LAAO) is an alternative to anticoagulation in the prevention of stroke in patients with non-valvular atrial fibrillation, especially in patients with a contra-indication for oral anticoagulation therapy (OAT).
Objective
Obtain long-term patient outcomes after successful LAAO in everyday clinical practice.
Methods
In this single-center registry spanning over 10 years, data of all consecutive patients that underwent percutaneous LAAO were collected. Observed thrombo-embolic and major bleeding event-rates after successful LAAO during follow-up were compared to expected event-rates based on CHA2DS2-VASc and HAS-BLED scores. Furthermore, anticoagulation and antiplatelet use during follow-up was evaluated.
Results
Of 230 patients scheduled for LAAO (38% female, aged 69.5±8.2 years, CHA2DS2-VASc 3.9±1.6, HAS-BLED 2.9±1.0) 218 patients had a successful implant (95%) with a follow-up duration of 5.2±3.1 years. The procedure was combined with catheter ablation in 52% of the patients. Fifty thrombo-embolic complications (24 ischemic stroke, 26 TIA) were observed during follow-up in 40/218 patients (18%). Ischemic strokes occurred with a rate of 2.1/100 patient-years, accounting for a 66% relative-risk reduction compared to the CHA2DS2-VASc predicted event-rate. Device-related thrombus was observed in 5 patients (2%). Sixty-five non-procedural major bleeding complications occurred in 24/218 patients (11%) with a rate of 5.7/100-patient years, which is comparable to estimated HAS-BLED-bleeding rates under OAT use. At end of follow-up 71% of all patients were on single antiplatelet or no antiplatelet/anticoagulation treatment, while 29% were on OAT.
Conclusions
Thrombo-embolic event-rates during long-term follow-up after successful LAAO remained consistently lower than expected supporting the efficacy of LAAO.
Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2021;42(5):373-498.
. Therefore, patients with AF and an increased stroke risk based on the CHA2DS2-VASc score are deemed to require oral anticoagulation therapy (OAT) to prevent for thrombo-embolic complications
Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2021;42(5):373-498.
. Oral anticoagulation therapy is highly effective in reducing the risk of stroke but has several disadvantages, including the increased risk of bleeding. A difficult tradeoff between competing risks frequently yields patients with sub-optimal treatment for the prevention of stroke, due to termination or ineffective dose reductions of OAT.
An alternative for stroke prevention in patients with non-valvular AF is percutaneous occlusion of the left atrial appendage (LAA), considered to be the main source (∼90%) of cardiac thrombi in non-valvular AF
. Left atrial appendage occlusion (LAAO) has been proven to be non-inferior compared to warfarin for stroke prevention in patients with non-valvular AF by two randomized controlled trials; the PROTECT-AF and PREVAIL
. In the combined 5-year outcomes of these trials the ischemic stroke or systemic embolism (SE) rate was numerically higher in the LAAO device arm, although not reaching statistical significance (HR 1.71, p=0.08). In contrast, differences in hemorrhagic stroke, death and non-procedural bleeding significantly favored LAAO (HR 0.20, p=<0.001; HR 0.73, p=0.04; HR 0.48, p= <0.001, respectively).
According to current guidelines LAAO may only be considered in patients with contra-indications for long-term OAT
Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2021;42(5):373-498.
, although this specific patient population has not been included in any published randomized controlled study to date. Nevertheless, large observational studies strongly suggest that LAAO is also very effective in patients that are non-eligible for OAT
Boersma L V, Ince H, Kische S, et al. Evaluating Real-World Clinical Outcomes in Atrial Fibrillation Patients Receiving the WATCHMAN Left Atrial Appendage Closure Technology: Final 2-Year Outcome Data of the EWOLUTION Trial Focusing on History of Stroke and Hemorrhage. Circ Arrhythmia Electrophysiol. 2019;12(4):1-13.
, long-term results on the efficacy of LAAO beyond 2 years, in patients with a contraindication for OAT, remain scarce. The main goal of this study was to determine if successful LAAO as an alternative for OAT maintains its effectiveness during long-term follow-up in everyday clinical practice.
Methods
Study design.
This single-center, non-randomized, prospective cohort study was designed to collect data of all consecutive patients scheduled for LAAO in the St. Antonius Hospital (Nieuwegein, the Netherlands). The inclusion period started directly after introduction of LAAO in 2009 and ended when a new generation LAAO device (WATCHMAN FLX, Boston Scientific, Natick, MA, USA) was implemented in 2019. All patients aged ≥18 years, scheduled for LAAO were included. The study was performed in accordance with the Declaration of Helsinki principles and with approval of the local ethics committee (MEC-U). Informed consent was obtained in accordance with local ethical guidelines. Study data were collected and managed using REDCap (Research Electronic Data Capture) electronic data capture tools.
Procedural characteristics – left atrial appendage occlusion.
All patients were implanted with LAAO devices that were CE–approved and available at the St. Antonius Hospital during this study (WATCHMAN 2.5, Boston Scientific, Natick, MA, USA and Amplatzer Amulet, Abbott, Minneapolis, MN, USA). Device type was chosen at the discretion of the implanting cardiologist. LAAO was performed according to local standard of care in agreement with consensus statement documents previously published
Tzikas A, Jr DRH, Gafoor S, et al. Percutaneous Left Atrial Appendage Occlusion: the Munich Consensus document on Definitions, Endpoints and data collection requirements for clinical studies. Eurointervention. 2016;(12):103-111.
. Successful LAAO was defined as a device deployed and implanted in correct position, meeting all release criteria according to the manufacturer’s instruction for use. In case of coexisting indications for other percutaneous cardiac procedures (e.g. catheter ablation (CA) for AF), LAAO was allowed to be combined and was in these cases performed directly afterwards, reducing the risk of multiple procedures
. All procedures were performed with trans-oesophageal (TOE) visual guidance. Prior to transseptal puncture, the left atrium, especially the LAA, was examined to exclude the presence of cardiac thrombi. Device sizing was performed during the procedure based on TOE imaging combined with fluoroscopy. Post-procedural antithrombotic regimen was not prespecified per protocol. Though aimed to comply to the manufacturer’s instruction for use, could be adjusted based on patient characteristics and after multidisciplinary consultation if needed.
Clinical follow-up
Post-implant visit was planned according to standard of care approximately three months after implantation and combined with LAA imaging (TOE or computed tomography(CT)) to assess device positioning, peri-device leakage (PDL) and device-related thrombus (DRT). Long-term follow-up data was obtained by inspection of electronic medical records. Additionally, patients were contacted by mail with short questionnaires to acquire data on clinical events and medication use during follow-up. If necessary, general practitioners and referring hospitals were consulted to complete the required follow-up information. Patients were censored at last follow-up date. The Dutch national registry for inhabitants was used to check if the included patients were alive or deceased at the end of study.
Outcomes
The goal of this study was to obtain patient outcomes after successful LAAO during long-term follow-up in everyday clinical practice, with a focus on stroke and non-procedural bleeding. The primary efficacy outcome of LAAO was the annualized event-rate of ischemic stroke. Secondary outcomes were; the composite annualized event-rate of ischemic stroke, SE and transient ischemic attack (TIA); annualized event-rate of major bleeding (Bleeding Academic Research Consortium (BARC) >2); time to first ischemic stroke and time to first major bleeding, the occurrence of DRT and death. All neurological thrombo-embolic events (TIA and ischemic stroke) were re-evaluated by a dedicated neurologist to improve data quality. Definitions of TIA and ischemic stroke are described in Supplementary table 1, all other outcomes are described in the earlier published LAA consensus statement documents
Tzikas A, Jr DRH, Gafoor S, et al. Percutaneous Left Atrial Appendage Occlusion: the Munich Consensus document on Definitions, Endpoints and data collection requirements for clinical studies. Eurointervention. 2016;(12):103-111.
. Events that occurred peri-procedural or within seven days post LAAO and related to LAAO as indicated by the treating physician, were considered peri-procedural events and were excluded from long-term event-rates because the main focus was on the long-term effect of LAAO. Events during follow-up that were the undoubted sequelae of other interventions (e.g. coronary artery bypass grafting, CA) were excluded. Lastly, anticoagulation and antiplatelet use during the study was evaluated at baseline (prior to LAAO), at discharge, at the first follow-up visit, after one year of follow-up and at end of study.
Statistical analysis
Continuous variables are presented as mean ± standard deviation or median [interquartile range] depending on the normality of the distribution of the variables. Categorical variables are depicted as numbers (percentages). Continuous variables are compared using the Student’s t-test (normal distribution) or Mann-Whitney U test (non-normal distribution) and categorical variables are compared using the χ2-test or Fisher’s Exact test. Event-rates are presented as number of events per 100 patient-years. Observed ischemic stroke event-rates are compared to the extrapolated expected event-rates from historic data
Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: The Swedish Atrial Fibrillation cohort study.
of patients not using any OAT, based on the mean CHA2DS2-VASc prior to LAAO of the studied cohort. The observed major bleeding event-rates are compared to the extrapolated expected event-rate from historic data
Comparative validation of a novel risk score for predicting bleeding risk in anticoagulated patients with atrial fibrillation: The HAS-BLED (hypertension, abnormal renal/liver function, stroke, bleeding history or predisposition, labile INR, elderly, drug.
of patients using OAT, based on mean HAS-BLED score prior to LAAO of the observed cohort. Kaplan-Meier curves are used to determine the incidence of outcomes over time. Log-rank test was used to compare Kaplan-Meier curves. Cox-regression proportional hazards analysis was used to adjust for possible confounding. Reported p-values are two sided and p-values <0.05 are considered statistically significant. Statistical analysis was performed using SPSS statistical software version 26.0 (SPSS Inc., Chicago, IL, USA) and R software version 1.3 (R Foundation for Statistical Computing).
Results
Patient and procedural characteristics
A total of 230 consecutive patients were scheduled for LAAO. Mean age of the population was 70±8 years. Included patients were at high risk for stroke and bleeding with a mean CHA2DS2-VASc score of 3.9±1.6. and mean HAS-BLED score of 2.9±1.0. In 109/230 (47%) patients standalone LAAO was performed, while 119/230 (52%) patients underwent LAAO combined with CA for AF and two patients underwent MitraClip implantation before LAAO (1%). The main target in all patients undergoing CA prior to LAAO was to achieve pulmonary vein isolation. In 9/230 patients (4%) LAA implantation was not attempted (n=4) or failed (n=5) due to an unfavorable LAA anatomy, while LAA thrombus or tamponade prevented LAAO in 3/230 patients. In 17/230 (7%) LAAO procedure was technically successful but one or more peri-procedural complications occurred. Most common procedural complications were pericardial effusion/tamponade (n=5) and access site (groin) bleeding/pseudo aneurysm (n=5). The baseline demographics and procedural characteristics for all patients, including differences between standalone LAAO and LAAO combined with CA, can be found in table 1 and table 2, respectively. Lastly, 2/230 patients (1%) withdrew informed consent for follow-up (figure 1). The mean follow-up duration of the successfully implanted patients was 5.2±3.1 years, yielding a total of 1134 patient-years.
Table 1Baseline characteristics
Total cohort N=230
LAAO N=111
LAAO combined with CA N=119
p-value
Age in years
69.5 ± 8.2
73.9 ± 6.1
65.5 ± 7.8
<0.001
Sex– female
88 (38.3)
40 (36.0)
48 (40%)
0.587
Indication for LAAO Thrombo-embolic event under OAT Prior bleeding/high bleeding risk Miscellaneous
† After introduction of pediatric probes for transoesophageal echocardiographic guidance only sedation was required. ‡ Second device successfully implanted. § All air emboli presented as a short duration of ST elevation during procedure rapidly resolving without clinical consequences/ischemic complications. ¶ Two minor strokes and one TIA occurred. # Defined as a device deployed and implanted in correct position, meeting all release criteria according to the manufacturers’ instruction for use. LAAO=left atrial appendage occlusion, CA= catheter ablation, PE= pericardial effusion., TIA=transient ischemic attack, TSP=transseptal puncture.
During follow-up, a total of 24 ischemic strokes occurred in 22/216 patients (10%). The annualized event-rate of ischemic stroke was 2.1. Comparing the observed event-rate to the expected annual event-rate of 6.2 based on the CHA2DS2-VASc score, a relative-risk reduction (RRR) of 66% was observed. Additionally, 26 TIAs occurred in 18/216 (8%) patients and no SEs were observed. Relative-risk reductions, Kaplan-Meier curve and event-rates per year are depicted in figure 2A-2C. First ischemic stroke occurred while patients were using: OAT in 4/22 (18%), single antiplatelet (APT) in 15/22 (68%), and no antithrombotic medication in 3/22 patients (14%). OAT was adjusted after ischemic stroke in 14/22 (66%); OAT was re-initiated in 7/14 (50%), clopidogrel was started in 1/14 (7%), acetylsalicylic acid (ASA) was switched to clopidogrel in 4/14 (29%), ASA was added to clopidogrel in 1/14 (7%) and persantin was added to clopidogrel in 1/14 (7%), respectively. Recurrent stroke occurred in two patients; one patient continued ASA after the index stroke and switched to clopidogrel after recurrent stroke. The other patient had a fatal stroke while using dual APT after the index ischemic stroke.
Figure 2Thrombo-embolic and major bleeding events during long-term follow-up. (A) Observed thrombo-embolic event-rates compared to expected event-rates without any anticoagulation therapy based on the CHA2DS2-VASc score. (B) Kaplan-Meier curve time-to first ischemic stroke (C) Annualized ischemic-stroke rate per follow-up year. (D) Observed major bleeding event rates compared to expected event-rates under anticoagulation use based on the HAS-BLED score. (E) Kaplan-Meier curve major bleeding for the full cohort. (F) Annualized major bleeding rate per follow-up year. HHT= hereditary hemorrhagic telangiectasia, RRR= Relative-risk reduction, SE= Systemic embolism, TIA= Transient ischemic attack.
Cardiac follow-up imaging was available in 207/216 (96%) patients. In 3/207 (1%) patients DRT was discovered during routine follow-up while on OAT, and DRT was not directly associated with thrombo-embolic complications. All three patients continued OAT, in 2/3 patients DRT had resolved on follow-up imaging and the other patient did not have any follow-up imaging after DRT discovery. One patient suffered a TIA, occurring approximately 10 months after DRT resolution while still using OAT. No imaging was performed at the time of TIA to evaluate if DRT had reoccurred. Additionally, two DRTs were discovered in patients presenting with ischemic stroke more than three years after device implantation while using single APT or no antithrombotic mediation at all. One of these patients eventually had a second fatal ischemic stroke on dual APT, two months after index ischemic stroke. No follow-up imaging evaluating the effect of OAT on DRT of the other patient was available. Four of 203 patients (2%) had significant PDL (table 2) and despite continuation of OAT three of these patients developed thrombo-embolic events (1 TIA, 3 ischemic strokes) during follow-up. However, no statistically significant differences were found in the occurrence of ischemic stroke comparing patients with any PDL during follow-up to patients without PDL (p=0.286).
Bleeding events
A total of 65 non-procedural major bleeding events occurred in 24/216 patients (11%). The annualized event-rate of major bleeding was 5.7. The observed event-rate is comparable to the expected annual event-rate of 6.0 based on the HAS-BLED score. The high bleeding rate was mainly driven by 45 recurring events (69% of all bleeding events; 40 nose bleeds, 4 gastro-intestinal bleeds, 1 urinary tract bleed) resulting in severe anemia requiring blood transfusion in 8/216 patients (4%) that suffered from hereditary hemorrhagic telangiectasia (HHT). Hereditary hemorrhagic telangiectasia is a genetic disorder causing abnormal vessel formation leading to an extremely high risk for recurrent bleeding. Subgroup analysis excluding HHT patients resulted in only 20 major bleeding events in 16/204 patients (8%). Comparing the annualized event-rate of 1.8 in this cohort excluding the HHT patients with the expected event-rate of 5.9 based on the HAS-BLED score, a RRR of 69% was observed. Relative-risk reductions, Kaplan-Meier curve and event-rates per year are presented in figure 2D-F. Intracranial hemorrhage (ICH) occurred in 7/216 patients (3%) of which 2/7 (29%) were still on OAT. Five out of seven (71%) patients had a history of ICH prior to LAAO.
LAAO combined with catheter ablation vs. standalone LAAO
All patients treated with a combined procedure had co-existing indications for LAAO and CA. Patients treated with standalone LAAO more often had a history of prior bleeding (90%), were older (74 vs. 65, p<0.001) and had a significantly higher CHA2DS2-VASc (4.3 vs. 3.6, p=0.001) and HAS-BLED (3.3 vs. 2.5, p<0.001) score than patients treated with LAAO combined with CA more frequent without a history of prior bleeding (64%). Relative-risk reductions of observed ischemic strokes compared to expected event-rates were equally present in both cohorts, however, more pronounced in the combined LAAO with CA cohort (74% vs. 52%, figure 3A). As depicted in figure 3B, a trend towards more ischemic strokes could be observed in the standalone LAAO group, which did not reach statistical significance (p=0.114) and when adjusted for CHA2DS2-VASc and age this trend was no longer present (HR 1.1, CI [0.4-2.9], p=0.852). Statistically significant less bleeding complications occurred in the LAAO combined with CA group (p<0.001), figure 3C, but this difference did not remain significant after adjustment for HAS-BLED score, age and prior bleeding as referral reason (HR 1.8, CI [0.6-5.2], p=0.306).
Figure 3Standalone LAAO vs. LAAO combined with CA. LAAO= left atrial appendage occlusion, LAA= left atrial appendage, FU=follow-up, RRR= relative risk reduction
Several subgroup analyses were performed to identify possible differences concerning the efficacy of LAAO in specific patient groups. As presented in table 3, RRRs for ischemic stroke, comparing observed event rates with expected event rates, were consistently observed in all patient groups. Relative-risk reductions for major bleeding were more pronounced in patients with a history of hemorrhagic stroke and in patients referred for LAAO because of thrombo-embolic events under OAT. In contrast, patients referred for LAAO indicated by previous bleeding/high risk of bleeding, showed a RR increase compared to historic data. As mentioned in the previous paragraph, these event-rates were mostly driven by the frequent recurrent bleeding events in patients with HHT.
Table 3Subgroup analysis relative-risk reductions ischemic stroke and major bleeding
No. patients
Follow-up years
CHA2DS2-VASc
Expected ischemic stroke
Observed ischemic stroke
Relative Risk Reduction
HAS-BLED
Expected major bleeding
Observed major bleeding
Relative Risk Reduction
Full cohort
216
1134
3.9
6.2
2.1
66%
2.9
6.0
5.7
5%
Full cohort without HHT
204
1101
3.9
6.2
2.0
68%
2.8
5.9
1.8
69%
Hx ischemic stroke
69
378
4.6
7.4
2.1
72%
2.8
5.9
3.7
37%
Hx hemorrhagic stroke
72
317
3.9
6.2
3.2
48%
3.3
6.7
1.9
72%
Indication LAAO bleeding/high risk of bleeding
132
532
3.9
6.2
3.0
52%
3.2
6.6
11.7
+77%
Indication LAAO thrombo-embolic event under OAT
41
238
4.6
7.4
2.1
72%
2.9
6.0
0.8
87%
Observed event-rates ischemic stroke compared to expected event-rates without any anticoagulation therapy based on the CHA2DS2-VASc score. Observed event-rates major bleeding compared to expected event-rates under OAC use based on the HAS-BLED score. HHT= hereditary hemorrhagic telangiectasia, Hx= history of, LAAO=left atrial appendage occlusion, OAT= oral anticoagulation therapy.
During long-term follow-up 51/216 patients (23%) died. Cardiovascular death was observed in 12/51 (24%), non-cardiovascular death in 28/51 (55%) and cause of death was unknown in 11 out of 51 (22%). The Kaplan-Meier curve of all-cause death is presented in Supplementary figure 1. Causes of cardiovascular death were: acute coronary syndrome 2/13 (15%), heart failure 3/13 (23%), ischemic stroke 1/13 (8%), intracranial hemorrhage 3/13 (23%), ventricular fibrillation 1/13 (8%), pulmonary embolism 1/13 (8%) and sudden cardiac death 2/13 (15%).
Anticoagulation and antiplatelet therapy
A schematic visual representation of OAT and APT during the course of the study is presented in figure 4. After the first follow-up visit in most patients the antithrombotic regimen could be reduced, but 46/210 (22%) were still on OAT. One year after LAAO the number of patients using OAT diminished further to 23/194 (12%). Reasons for continuing OAT were hybrid strategy in patient with extreme high thrombo-embolic risk, planned cardioversion/catheter ablation, the occurrence of thrombo-embolic events, PDL, and suspect DRT. Between one year of follow-up and the end of study, OAT was re-initiated in 36/194 (19%). In 17/36 patients (47%) thrombo-embolic events occurred during follow-up. In 10/36 (28%) recurrence of symptomatic AF was observed and frequent cardioversion and/or CA was performed. One of 36 (3%) patients preferred using OAT combined with LAAO. For the remaining 8/36 (22%) patients no justification for restarting OAT could be found.
Figure 4Anticoagulation and antiplatelet use during the study. * Combined with low molecular weight heparin (∼2 weeks). DAPT= dual antiplatelet therapy, (N)OAC = (non vitamin-K) oral anticoagulation, SAPT= single antiplatelet therapy.
Patients using OAT at the end of study compared to patients not using OAT were slightly younger (67 vs. 70, p=0.034), had a lower HAS-BLED (2.6 vs. 3.0, p=0.032) and less often had a history of intracranial hemorrhage, major bleeding, or high risk of bleeding (p<0.001). On the opposite, a history of ischemic stroke was more common (49% vs. 25%, p=0.001), patients more often used OAT prior to LAAO (86% vs. 44%, p<0.001) and were more often referred for LAAO because of thrombo-embolic events under OAT (35% vs. 12%, p<0.001). Furthermore, a combined LAAO with CA was more frequent in patients with OAT at the end of study (76% vs. 44%, p<0.001). Kaplan Meier curves during follow-up showed less bleeding events (p=0.026) and more thrombo-embolic events (p=0.001) in patients using OAT at end of study compared to patients not using OAT.
Discussion
This prospective registry describes one of the longest follow-up durations (mean 5.2±3.1 years) after successful LAAO in everyday clinical practice and therewith adds important data evaluating the significance of this preventive therapy as an alternative for OAT. An event-rate of 2.1 ischemic strokes per 100 patient-years, in a very high-risk population for developing stroke, supports the efficacy of LAAO during long-term follow-up. Combination of LAAO with CA for AF was frequently performed (52%) and occurrence of ischemic stroke and major bleeding was non-significantly different to the standalone LAAO group when adjusted for CHA2DS2-VASc and HAS-BLED scores. At the end of study 71% of all patients were not using OAT, illustrating this therapy to be a viable option for patients with competing risks of stroke and bleeding and the need to avoid long-term OAT. Nevertheless, an important observation is that LAAO in everyday clinical practice is also frequently combined with OAT due to re-initiation or continuation of anticoagulation for various reasons during longer follow-up. Thus, the premised benefit of LAAO, eliminating the need for OAT was not always achieved.
Ischemic stroke during long-term follow-up
Publications with outcomes beyond 2-years of follow-up after LAAO in everyday clinical practice are very limited and frequently have shorter follow-up durations and/or smaller sample sizes compared to our cohort. Albeit, the ischemic stroke rates were consistently low with comparable RRRs as shown in supplementary figure 213–17. The EWOLUTION study
Boersma L V, Ince H, Kische S, et al. Evaluating Real-World Clinical Outcomes in Atrial Fibrillation Patients Receiving the WATCHMAN Left Atrial Appendage Closure Technology: Final 2-Year Outcome Data of the EWOLUTION Trial Focusing on History of Stroke and Hemorrhage. Circ Arrhythmia Electrophysiol. 2019;12(4):1-13.
, an important registry of WATCHMAN LAAO (n=1020, mean age 73±9, CHA2DS2-VASc=4.5±1.6) did publish some long-term (2-year) outcomes, an ischemic stroke rate of 1.3 events per annum was observed. Our results are in line although RRR was slightly lower (81% vs. 66%) likely due to variations in patient selection. The difference in event-rates was probably not caused by the length of observation time, as event-rates remained stable over time. The two landmark randomized trials, PROTECT-AF and PREVAIL, comparing LAAO to OAT in the prevention for stroke, reported 5-year follow-up results
. Noteworthy, the RCTs only included patients that were eligible for OAT and therefore differed from patients referred for LAAO in everyday clinical practice. The event-rates of ischemic stroke in the device-arms of the PROTECT-AF and PREVAIL were 1.4 and 1.7, respectively. Likewise, the continued access protocol registries following the RCTs (CAPI and CAPII) reported ischemic stroke event-rates of 1.3 and 2.219. The RRRs of these studies were between 69-79% and comparable to our results
. To our knowledge, no long-term results on combined LAAO with CA with comparable follow-up durations to our registry are published so far, but the RRR of ischemic stroke after 1- follow-up presented by Wintgens et al. was comparable to our results (78% vs. 74%).
Bleeding during long-term follow-up
As major bleeding is one of the most frequent indications for LAAO, the occurrence of post-LAAO bleeding is an important outcome. In our cohort, non-procedural major bleeding events were comparable to the expected event-rate under anticoagulation use based on the HAS-BLED score. The lack of a reduction in bleeding events was mainly caused by a small cohort of HHT patients with repetitive nosebleeds, inherent to the disease even in the absence of OAT. Excluding these patients from the analysis resulted in a high RRR of 69%. Of note, some patients with important chronic bleeding conditions will likely develop new bleeding events even without any OAT, although they will be better protected for thrombo-embolic events by LAAO. Withholding OAT in those patients is mainly desired to reduce the bleeding burden, but at the risk of undertreatment regarding thrombo-embolic prevention.
Combination of LAAO and CA
LAAO combined with CA has previously been presented to be feasible, safe and effective
. More than half of the patients in our cohort had a co-existing indication for rhythm control and therefore underwent LAAO combined with CA further establishing the feasibility of combined procedures. In our hospital LAAO is performed by electrophysiologists and is a straightforward strategy avoiding exposure of patients to multiple procedures. As expected, baseline characteristics differed between patients with LAAO standalone (appearing to be more ‘frail’) and LAAO combined with CA patients. No significant differences between LAAO standalone and LAAO combined with CA were found regarding ischemic stroke and major bleeding events when adjusted for confounders (e.g. CHA2DS2-VASc and HAS-BLED). A higher RRR (74% vs. 52%) of ischemic stroke was observed in patients treated with combined LAAO with CA compared to standalone LAAO. Reducing AF burden by CA may have contributed to the stroke reduction, although this concept has not been proven in previous literature. Of note, the study was not powered for these comparisons which should therefore be interpreted with caution.
Challenges of anticoagulation use in everyday clinical practice
The post-LAAO antithrombotic drug regimen to prevent DRT during endothelialization of the device is frequently discussed as an important problem in patients with high bleeding risk. In our cohort major-bleeding events did not occur predominantly during early follow-up, which may reflect the ability to tailor and individualize post-LAAO drug regimen based on patient characteristics. Nevertheless, further exploration of optimal post-LAAO drug treatments remains important, since bleeding events are likely to diminish with less intensive drug treatment and the development of DRT and the effect of post-LAAO drug treatment is still poorly understood. Five of 207 patients with follow-up imaging available developed DRT in our cohort (2%), which is comparable to earlier published studies
. However, the presence of DRT in our cohort may have been underestimated since imaging for DRT was typically performed only once in our cohort, while repeated imaging may yield new DRT despite a clean TOE/CT after 3 months. Interestingly, even after the occurrence of thrombo-embolic events additional cardiac imaging was often omitted by treating neurologists and/or cardiologists, which may be due to lack of awareness of the DRT issue. In two patients DRT was only discovered after thrombo-embolic events occurred very late after LAAO, while three patients developed DRT during OAT. These extreme events stress the importance of developing a better understanding of DRT mechanisms and their relation to stroke, and optimal anticoagulation.
Another important observation during our study was that some patients continued OAT use and there was even an increase of OAT use between year 1 of follow-up and end of study. In patients with a very high-risk for developing recurrent thrombo-embolic events, sometimes a hybrid strategy of LAAO combined with OAT was chosen by their treating cardiologist, despite a lack of evidence for this strategy. The combined strategy of LAAO on top of OAT in patients with very high risk of thrombo-embolic events needs further investigation. Several patients restarted OAT because of new thrombo-embolic events during follow-up. Although re-initiation may seem reasonable and good clinical practice in these cases, there is no evidence that these patients will benefit from treatment with OAT. Patients using OAT also develop recurrent events, and similar to the ineffectiveness of changing OAT regimen
, it is questionable whether this will be the case in patients treated with LAAO. Additionally, the initial reason for avoiding OAT most likely still exists, and re-initiation may therefore even be harmful. Remarkably, there was a high proportion of patients in whom OAT was re-initiated without any reason or because of planned cardioversion/ablation and omitted to be discontinued afterwards. Since LAAO is not reimbursed in the Netherlands the concept of LAAO may not be familiar and LAAO may not be embraced by all physicians, possibly contributing to redundant OAT use in post LAAO patients.
Limitations
Our study was meant to show long-term outcomes of successful LAAO in everyday-clinical practice without a strict pre-specified protocol. The study lacks an active control group for comparison and the population is heterogeneous because of limited inclusion criteria. Despite historic data frequently being used to predict expected event-rates based on CHA2DS2-VASc and HAS-BLED scores to compare with observed event-rates in patients after LAAO, direct comparison should be done with caution. Expected event-rates used in this study were based on the most common used historic data in LAAO publications
Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: The Swedish Atrial Fibrillation cohort study.
. This underscores the challenge of estimating the actual treatment effect of LAAO without an active control group. Expected event-rates were calculated using baseline HAS-BLED and CHA2DS2-VASc. Changes during long-term follow up were not adjusted for. The use of any APT or OAT after LAAO also lead to stroke-risk reduction and bleeding risk increase which cannot be adequately adjusted for. Additionally, adequate AF burden information was not collected, possible influences of catheter ablation (52% of the patients) on stroke risk reduction may be present. Another limitation of this study is the lack of strict pre-specified follow-up visits. At the end of study patients were sent questionnaires to complete missing data, but recall bias may have resulted in lower event-rates. Lastly, other (newer generation) LAAO devices have appeared on the market since this cohort. Consequently, the results may not be extrapolated to all percutaneous LAAO devices and strategies.
Future perspectives
Unfortunately obtaining randomized controlled data in patients with a contra-indication for OAT appears to be difficult, illustrated by the early termination of the ASAP-TOO trial (Clinical trials.gov: NCT02928497), probably due to the sense of withholding a patient from appropriate treatment when randomized to the control group by many physicians. Another attempt for gaining randomized controlled evidence, the COMPARE-LAAO (Clinical trials.gov: NCT04676880), is currently enrolling patients in the Netherlands.
Conclusion
In conclusion, successful LAAO appears to be effective at preventing stroke during very long-term follow up cohort describing everyday clinical practice (RRR 66%). These results strengthen the position of LAAO as an option to avoid OAT, nevertheless, randomized controlled trial information in patients non-eligible for long-term OAT use are still highly anticipated. Of note, the premise of LAAO eliminating the need for OAT was not always achieved, which is an important observation for patient counseling and consideration of LAAO in everyday clinical practice.
Ethics Statement
The study was performed in accordance with the Declaration of Helsinki principles and reviewed and approved by the Institutional Review Board (IRB).
Mesnier J, Cruz-gonzález I, Peral V, et al. Ten-Year Outcomes Following Percutaneous Left Atrial Appendage Closure in Patients With Atrial Fibrillation and Absolute or Relative Contraindications to Chronic Anticoagulation. Circ Cardiovasc Interv. 2021;(August):1-3.
Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio-Thoracic Surgery (EACTS). Eur Heart J. 2021;42(5):373-498.
Boersma L V, Ince H, Kische S, et al. Evaluating Real-World Clinical Outcomes in Atrial Fibrillation Patients Receiving the WATCHMAN Left Atrial Appendage Closure Technology: Final 2-Year Outcome Data of the EWOLUTION Trial Focusing on History of Stroke and Hemorrhage. Circ Arrhythmia Electrophysiol. 2019;12(4):1-13.
Tzikas A, Jr DRH, Gafoor S, et al. Percutaneous Left Atrial Appendage Occlusion: the Munich Consensus document on Definitions, Endpoints and data collection requirements for clinical studies. Eurointervention. 2016;(12):103-111.
Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: The Swedish Atrial Fibrillation cohort study.
Comparative validation of a novel risk score for predicting bleeding risk in anticoagulated patients with atrial fibrillation: The HAS-BLED (hypertension, abnormal renal/liver function, stroke, bleeding history or predisposition, labile INR, elderly, drug.
Mesnier J, Cruz-gonzález I, Peral V, et al. Ten-Year Outcomes Following Percutaneous Left Atrial Appendage Closure in Patients With Atrial Fibrillation and Absolute or Relative Contraindications to Chronic Anticoagulation. Circ Cardiovasc Interv. 2021;(August):1-3.
3 St. Antonius Hospital, Neurology Department, Koekoekslaan 1, 3455CM, Nieuwegein, the Netherlands
Brief running title: Long-term Outcomes of Left Atrial Appendage Occlusion
Conflict of interest statement:
M. Maarse was supported by an educational grant from Boston Scientific. L. Boersma is a consultant for Boston Scientific and proctor for Abbott. M. Swaans reports proctoring fees for training/educational services to the Department of Cardiology from Boston Scientific, and personal fees from Abbott. The other authors declared no conflict of interests.
Key findings
This prospective registry describes one of the longest published follow-up durations after left atrial appendage occlusion (LAAO) in everyday clinical practice. An event-rate of 2.1 ischemic strokes per 100 patient-years, in a high-risk population for developing stroke, support successful LAAO remaining effective during long-term follow-up.
Combination of LAAO with catheter ablation for AF was frequently performed (52%). The occurrence of ischemic stroke and major bleeding during follow-up was non significantly different to the standalone LAAO group when adjusted for confounders.
Most common indication for LAAO based on this cohort is previous bleeding. At the end of study 71% of all patients were not using oral anticoagulation, illustrating that LAAO is a viable option for patients with competing risks of stroke and bleeding and the need to avoid long-term anticoagulation.
LAAO was also observed to be combined with anticoagulation for various reasons. This observation raises questions and should be further evaluated.
00063-6&id=gr1.jpg){kind=link}
00063-6&id=gr2.jpg){kind=link}
00063-6&id=gr3.jpg){kind=link}
00063-6&id=gr4.jpg){kind=link}