The present single-center, observational, prospective, and nonrandomized study included a consecutive cohort of patients who underwent LAAC using the LAmbre device (Lifetech Scientific Co, China) and were discharged with SAPT.

The inclusion criteria were age ≥ 18 years, diagnosis of nonvalvular AF with indication for indefinite OACs, and indication for LAAC due to absolute or relative contraindication to adequate doses of prolonged anticoagulants (approved by an expert committee based on patient history and current clinical practice guidelines). All consecutive patients between May 2017 and March 2021 were enrolled. The exclusion criteria were suboptimal device implantation and need for DAPT or OACs for other medical reasons (such as percutaneous coronary or vascular revascularization in the last 12 months). All patients underwent LAAC with the LAmbre device during the study recruitment period. During this period, no other devices were used for LAAC. Figure 1 describes the population included in the study.

Figure 1.

Flowchart of patients included in the study. DAPT, dual antiplatelet therapy; (D)OACs, (direct) oral anticoagulants; LAAC, left atrial appendage closure; p, patients; PTA, percutaneous transluminal angioplasty; SAPT, single antiplatelet therapy.

(0.18MB).

The procedures were performed by interventional cardiologists who were experts in LAAC with the support of an anesthesia and cardiac imaging team. Transesophageal echocardiography (TEE) or intracardiac echocardiography (8%) was conducted in all patients to rule out the presence of left atrial appendage thrombosis and guide the procedure. At 24hours after the procedure, transthoracic echocardiography was performed to ensure correct device placement and exclude pericardial effusion. If no complications occurred, patients were discharged at 24hours with SAPT (aspirin, clopidogrel, or triflusal).

The study protocol was approved by the Ethics Committee for Medical Research of the institution (CEIm-PSMAR). All patients signed an informed consent form before the procedure. All study procedures were conducted in accordance with the principles of the Declaration of Helsinki and ISO 14155 guidelines.

We collected baseline clinical data (sociodemographic data, medical history, and ischemic and bleeding risk according to the CHA2DS2-VASc16 and HAS-BLED17 scales), blood test data (renal function and complete blood counts), and echocardiographic data (measurement of the left atrial appendage and ventricular function and determination of valvular heart diseases). Technical data were collected during the LAAC (size of device used, procedural time, size of left atrial appendage on fluoroscopy, and procedural success), as well as clinical data (length of hospital stay and local and in-hospital complications).

In-person clinical follow-up was conducted at 3 and 12 months after the LAAC. All patients were treated with SAPT for at least 1 month after the procedure. SAPT discontinuation was at the discretion of the treating cardiologist based on relevant medical history (eg, ischemic heart disease, peripheral vascular disease), the patient's bleeding risk (based on previous events and their severity, as well as HAS-BLED score), and TEE data during follow-up (presence of residual leaks).

Device monitoring was conducted using TEE imaging at 1 and 12 months after the LAAC. The presence of periprosthetic leaks and their size were recorded, as well as images indicating DRT. DRTs were confirmed using synchronized cardiac computed tomography (CT). If thrombosis was found, it was directly reported to the team for early assessment and treatment.

Primary endpoints were defined as a composite of ischemic events during follow-up: stroke, transient ischemic attack, systemic embolism, and DRT. Cardiovascular and all-cause mortality were also recorded.

Secondary endpoints were the presence of bleeding events during follow-up. Bleeding events were classified according to the mean Bleeding Research Academy Consortium (BARC) score,18 and major bleeding was defined as any event catalogued as BARC ≥3a (3a: overt bleeding with a hemoglobin drop >3g/dL or need for transfusion with overt bleeding; 3b, overt bleeding with a hemoglobin drop >5g/dL or need for surgery/vasopressors; 3c, intraocular or intracranial bleeding; 5, fatal bleeding).

Statistical analysis was performed with IBM SPSS Statistics version 25. Continuous variables are expressed as mean ± standard deviation or median [interquartile range] according to distribution normality. Categorical variables are expressed as frequencies and percentages. Events during follow-up are expressed as the percentage of the total and as the event rate per 100 patient-years. The predicted rate of ischemic stroke was calculated from the CHA2DS2-VASc score adjusted for aspirin.19 The predicted rate of major bleeding was calculated according to the overall cohort of the HAS-BLED risk score.17 The relative risk reduction was calculated as (estimated annual rate %−observed annual rate %) / estimated annual rate %.

Overall, 86 patients underwent LAAC with the LAmbre device between May 2017 and March 2021; of these, 74 were enrolled in the study. Thus, 12 patients were excluded; the reasons for their exclusion are summarized in figure 1.

The baseline data and medical history of the patients included in the study are summarized in table 1. The population had a median age of 77 [72-83] years and a high prevalence of comorbidities (hypertension, 92%; type 2 diabetes mellitus, 49%; chronic kidney disease, 72%) and very high ischemic and bleeding risks (CHA2DS2-VASc, 4 [3-6]; HAS- BLED 4 [4-5]).

In total, 81% of patients had already had at least 1 major bleeding event; the most frequent origin was gastrointestinal (66.7%), followed by hemorrhagic stroke (20%). In addition, 31% of the patients had already experienced 1 cerebral bleeding event (stroke/transient ischemic attack). Three patients (4%) developed an acute complication during the procedure: 2 had femoral hematoma (not requiring surgical intervention) and 1 had cardiac tamponade requiring urgent pericardiocentesis that resolved. All patients were discharged after the procedure with SAPT (92% with aspirin).

TEE-mediated follow-up is summarized in table 2. TEE was obtained at 1 month of follow-up after the LAAC in 90.5% of patients. All patients had a technically optimal outcome after the closure, with a 0% rate of significant periprosthetic leaks (≥5mm). An image suggesting DRT was obtained from 1 patient (1.5%); this was confirmed with cardiac CT. This patient did not experience ischemic or embolic events during follow-up.

Follow-up TEE was performed at 12 months in 70.2% of patients; none had significant periprosthetic leaks while 2 (3.85%) had images compatible with DRT. Neither of these patients experienced ischemic or embolic events before or after their diagnosis. The patients with DRT were treated with low-dose OACs (apixaban 2.5mg/12h) until the leak resolved on cardiac CT.

All patients completed follow-up (table 3). The median follow-up duration was 2.5 [1.65-3.38] years, giving a total of 188 patient-years. SAPT duration was a median of 6 [3-30] months. One patient (1.3%) developed an ischemic stroke during follow-up, giving an annual incidence of 0.5%: this event occurred 2 years after the LAAC. Given the expected incidence of stroke according to CHA2DS2-VASc score in our embolic events during follow-up. cohort of 5.5%/y,19 this represents a relative risk reduction of 87.5% (figure 2).

Figure 2.

Predicted and observed rates of stroke and major bleeding with the use of single antiplatelet therapy after LAAC. BARC, Bleeding Academic Research Consortium; LAAC, left atrial appendage closure.

(0.13MB).

Regarding the secondary endpoints, 12 patients (16.2%; annual incidence, 6.4%) developed major bleeding during follow-up (BARC ≥ 3a). The theoretical incidence of bleeding according to the median HAS-BLED score was 8.7%/y, which represents a relative risk reduction of 26.4% vs the actual incidence of bleeding events (figure 2). There were no fatal bleeding events or bleeding requiring urgent surgical or endoscopic intervention. Notably, 10 of these 12 patients (83%) had a history of bleeding and 80% of these 10 patients had repeated bleeding of the same origin (gastrointestinal). In addition, bleeding events were associated with high mortality during follow-up (41.6%), with a mean of 9 months until death after the bleeding event. Bleeding events are described in greater detail in table 4. Two patients (2.7%) died from cardiovascular causes; neither was related to the device. In addition, 19 patients (25.7%) died from other causes during follow-up, a median of 2 [1-2.6] years after the LAAC. Noncardiovascular deaths are analyzed in detail in table 1 of the supplementary data (figure 3).

Figure 3.

Central illustration. Graphical summary of the study. LAAC, left atrial appendage closure; RRR, relative risk reduction; SAPT, single antiplatelet therapy; TIA, transient ischemic attack.

(0.45MB).

Antithrombotic therapy after LAAC is prescribed to avoid DRT and, consequently, the onset of new systemic embolic events. However, the biological process of intracardiac device endothelialization is not completely understood and the time required to complete the endothelialization process is unknown.20,21

Both the incidence and clinical consequences (embolisms) of DRT after LAAC vary widely among the published series. The most recent studies report annual incidences of 0.9% to 7.2%.21 The largest European registry, published by the EWOLUTION researchers with the WATCHMAN device,12 found a DRT incidence (identified by TEE or CT) of 4.1% (34 of 835 patients); 91% of cases were detected before 90 days after implantation (median, 54 days). None of the patients with DRT developed thromboembolic events such as stroke/transient ischemic attack or peripheral embolism during a 21-month follow-up. Similarly, in the Amplatzer Cardiac Plug retrospective multicenter study,22 the incidence of DRT (diagnosed by TEE) was 3.2% and was not associated with thromboembolic events.

This contrasts with the study published by Dukkipati et al.,23 in which the authors retrospectively analyzed the patients included in the PROTECT-AF, PREVAIL, CAP, and CAP2 pilot studies. All cases of DRT were diagnosed using TEE with high protocol adherence (90.9%-98.7%). Among a total of 1739 patients, DRT was found in 65 (3.74%). The proportion of systemic embolism was higher in the DRT group: 6.28 vs 1.65/100 patient-years (P<.001). Multivariate analysis revealed the following predictors of DRT: history of stroke/transient ischemic attack, permanent AF, left atrial appendage diameter, and left ventricular ejection fraction. Along these lines are the study results obtained by Simard et al.24 on predictive factors for DRT. The authors of the study performed a matched case-control analysis with and without DRT in a 1:2 proportion with various LAAC devices. DRT was associated with a higher incidence of major cardiovascular events during follow-up (29.5% vs 14.4%; P<.001), particularly ischemic stroke (16.9% vs 3.6%; P=.01). No effect was found for the antiplatelet therapy regimen prescribed after LAAC (SAPT vs DAPT) on DRT onset. Two main predictors of DRT were identified—hypercoagulability and the presence of pericardial effusion—and several minor predictors: renal failure, implantation depth >10mm from the pulmonary veins, and nonparoxysmal AF. The presence of 1 major risk factor or 2 minor factors conferred a 2.1-fold increased risk of DRT during follow-up.

Regarding the LAmbre device, very low rates of DRT have been reported during follow-up, between 0% and 1.3% at 12-months of follow-up with standard DAPT (for at least 3 months).10,25,26 Our study is the first to show DRT data for this type of device with the use of SAPT.

In our study, the DRT incidence was 1.5% at 30 days after LAAC and 3.85% at 12 months, which represented 4% of the total cohort and 2.5% per TEE performed. Although this incidence of DRT is higher than that previously described with the LAmbre device, these studies were performed in very small populations (n=11-24),26,27 with DRT follow-up times of up to 3 months,10 or with low rates of echocardiographic monitoring at 12 months of follow-up (60%).25

In general, we believe that this result is promising, considering that all of the above-mentioned DRT studies were performed with OACs or, typically, DAPT. The low incidence of DRT in our study could be explained by chance, given the relatively small cohort; however, we believe that the design of the LAmbre device and the material used, together with the implantation technique, facilitated a complete closure of the appendage and that this can promote the rapid endothelialization of the device.25,28Table 5 summarizes the main results for DRT from the above-mentioned studies.

In the present study, the observed rate of stroke was 0.5%/y, which indicates a relative risk reduction of 91% vs the expected rate of events according to the CHA2DS2-VASc score. This is a highly pertinent result because, first, it shows that the rate of ischemic cerebral events in patients with SAPT after LAAC is extremely low and, second, it is comparable with the results of other studies that used more aggressive antithrombotic therapy regimens after LAAC. In the EWOLUTION study12 (which included 1020 patients with LAAC treated with a WATCHMAN device and had a 2-year follow-up), 60% of patients received DAPT after LAAC and 27% were treated with OACs for at least 2 months. During follow-up, the annual rate of ischemic stroke was 1.3% and the risk reduction was 83%. In addition, the European Amplatzer registry29 included 1047 patients followed for an average of 13 months. Overall, 59% of the patients were discharged with OACs or DAPT after the LAAC (mean duration, 3.8 months). In this case, the annual rate of stroke/transient ischemic attack was reported to be 2.3% and the risk reduction was 59%. Finally, if we consider the initial descriptive study of the LAmbre device in Europe, Park et al.25 reported a rate of cerebral ischemic events of 2%/y with a 3-month duration of DAPT.

As far as we know, only 1 study has assessed the use of SAPT after LAAC with the ACP and Amulet devices in a population with high bleeding risk.30 In that nonrandomized study, Korsholm et al. included 107 patients who were prescribed aspirin monotherapy after the procedure. Both the study population and the results are comparable to those of our cohort. The cohort had a mean CHA2DS2-VASc score of 4.4±1.6 and a HAS-BLED of 4.1±1.1. The reported stroke incidence was 2.3%/y during follow-up while the relative risk reduction was 61%.

Despite the reduction in antithrombotic therapy after LAAC, we still detected a high rate of major bleeding events during follow-up (12 patients, 16.2%; 6.4%/y). Because 3 patients were not taking SAPT at the time of the event (25% of bleeding events), their events cannot be considered related. This strengthens the theory that the population exhibits very high bleeding risk and a propensity for bleeding events, even without any antithrombotic therapy. However, the rates are clearly lower if the bleeding events in our population are compared with the theoretical bleeding risk according to the HAS-BLED scale. The theoretical incidence of bleeding according to HAS-BLED score would be 8.7%/y, which represents a relative risk reduction of 26.4% vs the actual incidence of bleeding events.

Our results agree with those of other studies analyzing the presence of bleeding events in real-life clinical practice and with shorter DAPT regimens after LAAC. In the above-mentioned 2-year subanalysis of the EWOLUTION study,12 a 2.7%/y rate of bleeding events was found in the total cohort of 1020 patients, with a relative risk reduction vs bleeding expected with HAS-BLED of 46%. If the subgroup of patients with bleeding history is analyzed, the incidence of events increases to 4.5%/y and the relative risk reduction is 30%.

The only other study to assess the SAPT strategy, by Korsholm et al.,30 found a major bleeding rate of 3.8%/y, with a relative risk reduction of 57% vs that expected with HAS-BLED. Although the population of that study is comparable to our population in terms of bleeding risk by HAS-BLED (mean, 4.1), the populations differ regarding other comorbidities, such as chronic kidney disease (13% vs 72% in our registry), which are directly related to the platelet dysfunction and altered hemostasis.32 This could explain our more discrete reduction in major bleeding vs Korsholm et al.30

The main limitations of the present study are due to its observational and nonrandomized design. The results are limited by the lack of a control group or alternative treatment validating the findings. In the absence of a control group, event prediction estimates of the CHA2DS2-VASc and HAS-BLED scales were used as an approximation. In addition, in our study, the treating physicians established the SAPT duration; studies are required to determine its optimal duration. Moreover, the sample size and TEE follow-up rate limit the interpretation of the results. The study was performed with a single device. Finally, the sample was drawn from a single center and its size is relatively small. Larger series and randomized experimental studies are required to confirm these promising findings.