You are here

  • Home
  • Online First
  • Left atrial appendage closure for patients with atrial fibrillation at high intracranial haemorrhagic risk

Article Text

Article menu
Original research
Left atrial appendage closure for patients with atrial fibrillation at high intracranial haemorrhagic risk
  1. Avia Abramovitz Fouks1,
  2. Shadi Yaghi2,
  3. Elif Gokcal1,
  4. Alvin S Das1,3,
  5. Ofer Rotschild1,
  6. Scott B Silverman1,
  7. Aneesh B Singhal1,
  8. Jorge Romero4,
  9. Sunil Kapur4,
  10. Steven M Greenberg1,
  11. Mahmut Edip Gurol1
  1. 1Neurology, Massachussets General Hospital, Harvard Medical School, Boston, MA, USA
  2. 2Neurology, Brown University, Warren Alpert Medical School, Providence, RI, USA
  3. 3Neurology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA
  4. 4Division of Cardiovascular Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA, USA
  1. Correspondence to Dr Mahmut Edip Gurol; MEGUROL{at}mgh.harvard.edu

Abstract

Background and objectives Although left atrial appendage closure (LAAC) is performed in patients with non-valvular atrial fibrillation (NVAF) at increased risk of intracranial haemorrhage (ICH), outcome data are scarce. We assessed the detailed neurological indications for LAAC and outcomes after LAAC in high ICH risk patients.

Methods Study population included consecutive patients with NVAF who underwent LAAC in a single hospital network between January 2015 and October 2021 because of prior ICH or the presence of high ICH risk imaging markers on brain MRI (cerebral microbleeds (CMBs)). Primary safety and efficacy outcome measures were the occurrence of ICH and thromboembolic events, respectively, after LAAC.

Results Among 146 patients with NVAF who underwent LAAC for high ICH risk, 122 had a history of ICH, while 24 presented with high ICH risk imaging markers only. Mean age was 75.7±7.61, 42 (28.8%) were women. Mean CHA2DS2-VASc score was 5.23±1.52. Of 122 patients with ICH history, 58 (47.5%) had intraparenchymal haemorrhage (IPH), 40 (32.8%) had traumatic ICH (T-ICH) and 18 (14.7%) had non-traumatic subdural haemorrhage. Of 85 patients with brain MRIs including necessary sequences, 43 (50.6%) were related to cerebral amyloid angiopathy and 37 (43.5%) to hypertensive microangiopathy. While 70% of patients were discharged on oral anticoagulants (OAC), 92% were not taking OAC at 1 year. Over 2.12 years mean follow-up, one patient had recurrent non-traumatic IPH (incidence rate (IR) 0.32 per 100 patient-years), five had T-ICH (IR 1.61 per 100 patient-years) and six had an ischaemic stroke (IR 1.94 per 100 patient-years).

Conclusions Among patients with NVAF at high ICH risk, LAAC demonstrated a low risk of recurrent ICH or ischaemic stroke compared with previously published data. LAAC in high ICH risk populations should be considered in clinical practice per FDA approval and recent guidelines.

  • Atrial Fibrillation
  • Hemorrhage
  • Anticoagulants
  • Stroke

Data availability statement

Anonymized data not published within this article will be made available by reasonable request from a qualified investigator.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/svn-2024-003142

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Patients with a history of intracranial haemorrhage (ICH) or small vessel disease imaging markers are at increased risk of ICH, especially while using oral anticoagulants (OAC). Left atrial appendage closure (LAAC) was approved by the FDA in 2015 for patients with atrial fibrillation (AF) who cannot tolerate lifelong OAC but data on the efficacy and safety of LAAC in high ICH risk patients is still scarce.

    WHAT THIS STUDY ADDS

    • In this study, we found low spontaneous ICH incidence rate of 0.32 per 100 patient-years in a very high ICH risk AF patient population who underwent LAAC while the reduction in ischaemic stroke risk of 74.2% was in line with previous OAC/LAAC trials compared with the expected risk based on CHA2DS2-VASc. We also addressed the different ICH aetiologies based on neuroimaging analysis demonstrating the success of LAAC in a diverse population at high ICH risk.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Our study supports the view that LAAC is a safe and efficient stroke prevention method among patients with AF at high ICH risk and our results lend further support to the new AF guidelines that provided a higher level Class of Recommendation for LAAC in this patient population.

    Introduction

    Patients with a medical history notable for intracranial haemorrhage (ICH) or presenting high ICH risk imaging markers on brain MRI, such as cerebral microbleeds (CMBs) and cortical superficial siderosis (cSS), are at increased risk for an ICH.1 2 The risk is further increased by long-term use of oral anticoagulants (OAC).3–7 A clinical dilemma arises when these high ICH risk patients concurrently have non-valvular atrial fibrillation (NVAF), as ischaemic stroke prevention in this context has classically relied on life-long OAC based on risk score.8 Two recent randomised controlled trials (RCTs) attempted to investigate the outcomes of using OAC versus permanently avoiding them in patients who experienced ICH. However, these studies failed to establish non-inferiority of OAC use compared with avoidance, as OAC arms showed higher ICH recurrence rates and very high OAC-related ICH mortality.4 5 Such high ICH risk patients also include patients who have significant cerebral small vessel disease (cSVD), which is the main cause of ICH in middle-to older age adults. The presence of CMBs and cSS is well validated markers of ICH risk, including warfarin-related and DOAC-related ICH and could help determine the aetiology of ICH and to estimate the recurrent ICH rate.1 9 In 2015, the FDA approved left atrial appendage closure (LAAC) with Watchman device (Boston Scientific, St Paul, Minnesota) in patients who have a reason to seek an alternative to life-long OAC use, based on two RCTs that showed non-inferiority of LAAC compared with warfarin in preventing ischaemic stroke. These studies also showed over 80% decrease in the risk of ICH after LAAC compared with long-term OAC use.10–12 More recently, Amplatzer Amulet device (Abbott, Minneapolis, Minnesota) was approved by FDA based on an RCT that showed that this device was non-inferior to Watchman for stroke prevention and procedural features.13 Similar to the DOAC studies, these LAAC RCTs also excluded patients with a history of prior ICH as one arm involved long-term OAC use. The long-term safety and efficacy outcomes of LAAC in high ICH risk populations remain a matter of debate due to the lack of randomised data. We, therefore, aimed to evaluate the outcomes in patients with NVAF who received LAAC because of a past ICH or the presence of high ICH risk markers on MRI.

    Methods

    Study population

    In this retrospective, observational study with longitudinal follow-up, we conducted a comprehensive review of all patients with NVAF who underwent LAAC at the specific hospital network between 1 January 2015 and 31 October 2021. During this timeframe, 561 LAAC procedures were identified in this hospital system with a total of over 2200 beds. Patients who had the left atrial appendage (LAA) closed using surgical techniques were not included as these procedures have substantial differences in the actual intervention, postprocedure follow-up and postprocedure antithrombotic treatment. For every patient, the referral indication for LAAC was identified through in-depth electronic medical records (EMR) (review conducted by a neurologist and only patients referred due to high ICH risk were included). High ICH risk was defined as either prior history of ICH or the presence of high ICH risk imaging markers on MRI performed prior to LAAC, namely CMBs or cSS. Based on EMR and available imaging at time of index haemorrhagic event, ICHs were categorised into intraparenchymal haemorrhage (IPH), non- traumatic subdural haemorrhage (SDH), traumatic ICH (T-ICH), aneurysmal/non-aneurysmal convexity subarachnoid haemorrhage (SAH), intraventricular haemorrhage and undetermined if no sufficient data were available. Patients with high ICH risk imaging markers but no history of ICH were referred to as CMB-group for the purpose of this study, as all had CMBs while some also had other markers.

    Clinical data collection

    Baseline demographics and medical history including cardiovascular risk factors at time of LAAC and history of acute ischaemic stroke (AIS) were obtained from patient charts. Information regarding any ICH prior to the LAAC procedure was documented, including the type of haemorrhage and OAC treatment at that time. Cardiac imaging data before LAAC, at the time of LAAC and at 6 weeks follow-up visit, were collected. Procedural details including type of LAAC device, procedure success and periprocedural complications were collected. Periprocedural complications were defined as the occurrence of pericardial effusion/tamponade, vessel/cardiac perforation, device migration, major bleeding, stroke, death or any condition that required surgical or other intervention within 7 days of LAAC. Device-related thrombus and peri-device leak, including its size, were documented through review of cardiac imaging at all time points. All patients had at least two TEEs, one at the time of the LAAC procedure and one at 6 weeks. Interval history from LAAC to last relevant clinical follow-up visit was reviewed for the occurrence of AIS, symptomatic ICH including T-ICH, myocardial infarction, systemic embolism and major bleeding. In case of AIS or ICH, detailed review of EMR and imaging studies was conducted to establish the underlying cause of the event. Antithrombotic medication use was recorded at the time of discharge after LAAC as well as at 6 weeks, 3 months, 6 months, 12 months after discharge and during the last available relevant follow-up visit. All data were collected using strict definitions by a neurologist under the supervision of a senior vascular neurologist.

    Standard protocol approvals, registrations and patient consents

    This study was performed with the approval of and in accordance with the guidelines of the institutional review board (IRB) of the hospital. As this was a retrospective study, IRB waived the requirement for the informed consent. There were no photographs, videos or other information of any recognisable person.

    Neuroimaging analysis

    Brain MRIs that were performed before LAAC were reviewed by a neurologist (AAF) and vascular neurologist (MEG) for the presence of cSVD markers as defined by STandards for ReportIng Vascular changes on nEuroimaging (STRIVE) criteria.14 15 Haemorrhagic cSVD imaging markers reviewed were deep and lobar CMBs and focal and disseminated cSS.16 17 Non-haemorrhagic imaging markers included white matter hyperintensities (WMHs) graded using the Fazekas score and WMHs patterns (multispot WMH pattern and peri-basal ganglia WMH pattern), deep and lobar lacunes and enlarged perivascular spaces (EPVS) in the basal ganglia and in the centrum semiovale using a validated 4-point visual rating scale.18–21 Severe WMHs were defined as Fazekas score >2 and severe EPVS was defined as EPVS >2. The presence of haemorrhagic and non-haemorrhagic imaging markers including CMBs, cSS, severe WMH, severe EPVS and lacunes were recorded for each patient to estimate the burden of MRI markers of high ICH risk.22 Finally, the aetiologic classification of every IPH was performed. Patients with strictly deep IPH with or without deep CMBs were classified as hypertensive cSVD (HTN-cSVD) related. Diagnosis of possible and probable Cerebral Amyloid Angiopathy (CAA) was based on the new Boston criteria V.2.0.23 Patients with mixed-location ICH/CMBs and no cSS were presumed to have predominantly HTN-cSVD while patients with mixed location ICH/CMBs and the presence of cSS were categorised as having real mixed pathology consistent with HTN-cSVD and CAA concurrently.1 24 25 The same neuroimaging analysis was performed for patients in the CMB-group. Patients with non-aneurysmal convexity SAH were included in the neuroimaging analysis as this type of haemorrhage is well known to be part of the clinical manifestations of CAA.23

    Outcomes

    The primary safety outcome measure of this study was the occurrence of symptomatic ICH (both non-traumatic and traumatic) and the primary efficacy outcome was thromboembolic events (AIS and systemic embolism) during the follow-up period after LAAC. Secondary outcomes were the occurrence of device-related thrombus, peri-device leak of more than 5 mm, myocardial infarction, major bleeding and all-cause death during the follow-up period. Further definitions of these outcome measures are provided in the attached online supplemental file 1.

    Supplemental material

    Statistical analysis

    Categorical variables were reported in counts and corresponding percentages while continuous variables were reported as either mean±SD or median and IQR. Patient-years were calculated from the time of LAAC until the occurrence of event of interest or censoring. The predicted annual risks of thromboembolism and major bleeding were estimated based on CHA2DS2-VASc and HAS-BLED scores and corresponding incidence rates in historical controls.

    Data availability statement

    Anonymised data not published within this article will be made available by reasonable request from a qualified investigator.

    Results

    A total of 561 patients underwent endovascular closure of the LAA using either the Watchman/Watchman-FLX (n=145) (Boston Scientific, St Paul, MN) or Amplatzer Amulet (n=1), (Abbott, Minneapolis, Minnesota). One hundred and twenty-two patients had LAAC due to a history of prior ICH and 24 due to CMBs identified on brain MRI. Among patients without ICH, all had CMBs while 5 (20.8%) also had cSS (CMB-group). Overall, 146 patients met the inclusion criteria for this study.

    Baseline characteristics and vascular risk factors of the study population at time of LAAC are shown in table 1.

    View this table:
    Table 1

    Baseline patient characteristics (n=146)

    The average age of patients at time of LAAC was 75.66±7.61 years, and 42 (28.8%) were women. Median duration from NVAF diagnosis to LAAC was 3.03 years (IQR) 0.67–6.73). Prior to LAAC, 36 (24.7%) patients had a history of AIS. The mean CHA₂DS₂-VASc and HAS-BLED scores were 5.23±1.52 and 3.58±1.00, respectively. Based on EMR and available imaging studies, all haemorrhagic pathologies that led to LAAC in our study population were classified in detail according to the type of haemorrhage and its location (table 2). Six patients had more than one ICH. For 2 (1.6%) patients, data regarding the type and location of the haemorrhage were not available.

    View this table:
    Table 2

    Classification of haemorrhagic pathologies that led to LAAC in our study population

    IPH aetiology was determined based on clinical data extracted from the EMR and neuroimaging analysis as described in the Methods section. Of 58 patients with IPH and four non-aneurysmal convexity SAH patients (62 total), 31 (50.0%) were CAA related, 26 (41.9%) HTN-cSVD related and 2 (3.2%) were of mixed pathology (mixed location ICH/CMB with cSS). In the CMB-group, underlying cSVD was determined in 23 patients with available MRI source images. Overall, 12 patients in the CMB-group had imaging markers in a CAA distribution and 11 patients had HTN-cSVD distribution (table 3).

    View this table:
    Table 3

    Presumed aetiology of IPH and of haemorrhagic imaging markers (CMB-group)

    Further review of these markers in all patients revealed that all had at least two high ICH risk imaging markers (92% in addition to IPH for IPH-group and 100% for CMB-group) (table 4).

    View this table:
    Table 4

    Burden of MRI markers of high intraparenchymal haemorrhage risk

    Anticoagulant treatment at time of index ICH is summarised in figure 1. Of 113 ICHs included in this analysis, 88 (77.9%) were identified as OAC-related haemorrhages.

    Figure 1
    Figure 1

    OAC type at index haemorrhage based on major intracranial haemorrhage subtypes. DOAC, direct oral anticoagulant.

    During the procedure, 117 patients had the Watchman 2.5 device implanted, 28 had the Watchman-FLX and one had the Amplatzer Amulet device. Regarding peri-procedural complications, two patients suffered pericardial effusion. None of the patients had peri-procedural AIS, ICH, non-CNS bleeding, myocardial infarction, other unplanned surgical and/or endoscopic intervention(s) or death.

    For the primary safety outcome of ICH, during mean follow-up of 2.12 years, one patient had recurrent non-traumatic IPH (IR 0.32 per 100 patient-years), five had T-ICH (IR 1.61 per 100 patient-years).

    Non-traumatic right frontal IPH occurred 3.12 years after LAAC in an 82-year-old patient who had prior history of left frontal convexity SAH due to probable CAA and right middle cerebral artery ischaemic stroke, while using aspirin only. Five patients suffered T-ICH between 2.5 months to 51 months after LAAC, all while using aspirin only. Two of them had LAAC due to a history of past IPH (one lobar CAA-related IPH and one deep HTN-related IPH), two due to a history of T-ICH and the last patient due to the presence of high-risk ICH imaging markers in a HTN-cSVD distribution. There was no death related to the T-ICH in follow-up.

    For the primary efficacy outcome of thromboembolic events, six patients sustained AIS between 5 months and 26 months after LAAC, with an IR of 1.94 per 100 patient-years. This represents a reduction of 74.2% when compared with the expected AIS rate of 7.51 per 100 patient-years based on weighted CHA₂DS₂-VASc (figure 2). Two of these patients were referred to LAAC due to the presence of high-risk imaging markers (one for lobar CMBs and one for mixed location CMBs), two due to non-traumatic SDH, one due to T-ICH and one due to prior lobar CAA-related IPH. All six AIS appeared to be of cardioembolic origin with one having a competitive aetiology of air embolism shortly after scuba diving. All AIS happened on antiplatelet monotherapy. Four post-LAAC AIS events happened during the first year after LAAC but none in the first 6 months. There were no events of systemic embolism during follow-up.

    Figure 2
    Figure 2

    The predicted ischaemic stroke rate in our cohort population based on weighted CHA₂DS₂-VASc score compared with the observed ischaemic stroke rate after LAAC. LAAC, left atrial appendage closure; RRR, relative risk reduction.

    For the secondary outcomes, transoesophageal echocardiogram (TEE) performed 6 weeks postprocedure revealed device-related thrombus in two patients while they were still on OAC treatment. On follow-up TEEs, thrombi resolved, and patients were switched to antiplatelet-only therapy, they did not have any AIS or ICH. Forty patients (27.4%) had peri-device leak on TEE 6 weeks postprocedure, but none of them had a peri-device leak greater than 5 mm or needed further intervention for this matter. AIS was not different between patients who had any leak when compared with patients without any leak (p=0.66). Two patients had myocardial infarction during follow-up period and three patients had a major non-cerebral bleeding while using antiplatelet therapy only. During the follow-up period, 20 patients passed away. Of them, 10 were non-cardiovascular related deaths, 4 were due to cardiovascular causes and 6 of unknown cause. Five patients died during the first year after LAAC.

    Post-LAAC antithrombotic treatment was individualised to each of the patients’ perceived ischaemic and haemorrhagic risks by their neurologist and cardiologist and varied as described in detail in table 5. One hundred and two (69.9%) patients were discharged on OAC treatment. At 1-year follow-up, 111 patients (76%) were using antiplatelet therapy while 10 (6.8%) were still using OAC.

    View this table:
    Table 5

    Antithrombotic treatment after left atrial appendage closure (n=146)

    Discussion

    In this retrospective cohort study of 146 patients who had LAAC for the indication of high ICH risk and structured follow-up, we observed low rates of symptomatic ICH and thromboembolic events over a mean follow-up of 2.12 years. Our analysis incorporated comprehensive neuroimaging analysis of ischaemic and haemorrhagic cSVD imaging markers as well as the aetiology of the primary IPH to better understand the severity of ICH risk in each patient. Our results, when compared with the published literature of OAC use in ICH patients, suggest that LAAC is a feasible alternative stroke prevention method in high ICH risk population including patients with high haemorrhagic risk imaging markers and no prior ICH. In addition, our results also suggest that in depth neuroimaging analysis including high haemorrhage risk markers can guide decision-making regarding stroke prevention methods in all patients with AF who can receive a brain MRI. This hypothesis should be further studied in large RCTs of LAAC for high ICH risk.

    Two large RCTs have studied non-traumatic IPH recurrence rates in patients with prior non-traumatic IPH who were (re)started on DOAC or avoided them. The trials failed to show non-inferiority of restarting OAC over avoiding them. The on treatment analysis of both studies showed very high recurrent ICH incidence rates while on OAC treatment compared with avoidance of OAC, 8.25 versus 1.65 per 100 patient-years in “Effects of oral anticoagulation for atrial fibrillation after spontaneous ICH in the UK: a randomised, open-label, assessor-masked, pilot-phase non-inferiority trial” (SoSTART) and 4.3 versus 0 per 100 patient-years in “Apixaban versus no anticoagulation after anticoagulation-associated intracerebral haemorrhage in patients with AF in the Netherlands: a randomised, open-label, phase 2 trial” (APACHE-AF), respectively.4 5 In our study, recurrent non-traumatic IPH incidence rates after LAAC were low at 0.32 per 100 patient-years despite their high HAS-BLED score (mean 3.58) and high burden of MRI markers of high IPH risk (92% of the ICH population and 100% of CMB-group reviewed had≥2 imaging risk markers). These results support an important role for LAAC as an AIS prevention method in AF patients that circumvents the need for lifelong OAC use and reduces the elevated risk of fatal/disabling OAC-related ICH.

    In the general population over 55 years of age, annual IPH risk is 0.09%.26 The annual risk of recurrent IPH ranges from 1.6% to 10.4% among IPH survivors without the use of anticoagulant.1 Recurrent IPH rates vary according to the type of underlying cSVD (hypertensive cSVD vs CAA), CAA being invariably associated with the highest ICH rates (both first time and recurrent).1 Another very recent development was that the data safety monitoring board of the Edoxaban for Intra-Cranial Hemorrhage Survivors with Atrial Fibrillation (ENRICH AF) study that compared edoxaban to aspirin for NVAF patients who survived an ICH, stopped the enrollment of all lobar IPHs and convexity SAH. This decision was based on observations of unacceptably high risks of recurrent IPH among lobar IPH and convexity SAH patients assigned to the edoxaban arm of ENRICH AF.27 In our study, the only patient with a recurrent non-traumatic IPH had probable CAA but 42 CAA patients did not have any IPH during follow-up. Based on ENRICH AF data, it will be difficult to defend even the performance of RCTs that could assign CAA or lobar IPH patients to a long-term DOAC arm, therefore data from observational single arm studies like ours will be important to help with therapeutic uncertainty. Additionally, these results highlight the significance of distinguishing between different underlying cSVD subtypes and considering the burden of cSVD imaging markers for assessing the individual patient’s ICH risk. By carefully assessing a patient’s ICH risk alongside their ischaemic risk, healthcare professionals can make informed decisions regarding the optimal stroke prevention strategy. The online supplemental table 1 summarises all published studies that explore the outcomes of LAAC in patients with prior ICH or high ICH risk.28 29 Some of these studies were summarised in a review article from year 2020.30 Our study is the largest to provide a comprehensive breakdown of ICH types and aetiologies and has a relatively long follow-up period. Our results align with those of prior studies, supporting its reliability and validity in terms of ischaemic/haemorrhagic outcomes and contributing to the existing data to support the role of LAAC in stroke prevention in patients at high ICH risk.

    A multicentre registry of NVAF patients treated with warfarin and suffering ICH, reported 40% of haemorrhages, were traumatic and 30-day mortality following T-ICH was 25%.31 Interestingly, in our study, five out of six post-LAAC ICH events were traumatic, with no related death. Our study cohort is limited in size to draw general conclusions on this issue but underscores the value of LAAC in NVAF patients at high risk for fall or trauma as a stroke prevention method.

    In NVAF patients at high risk of haemorrhagic events, the primary goal is to prevent an AIS while not disproportionately increasing ICH risk.12 Based on the calculated weighted CHA₂DS₂-VASc of our study population, the estimated annualised IR for ischaemic stroke was 7.51% per year. In our study, NVAF patients who underwent LAAC exhibited an IR of 1.94%, resulting in a relative risk reduction of 74.2%.32 These results are in line with OAC-related reductions in AIS rates based on phase 3 RCTs of OACs.33 The rate of ischaemic stroke may even be further reduced with the newest generation devices, which have a lower rate of peri-device leaks thanks to more available sizes and lower DRT risk thanks to the new antithrombotic HEMOCOAT Technology as in WATCHMAN FLX PRO.34 Combination of our safety and efficacy results demonstrates the substantial overall benefit achieved through LAAC, as it effectively reduces the risk of both haemorrhagic and ischaemic stroke in a high ICH risk patient population. The ‘2023 Guidelines for the Diagnosis and Management of Atrial Fibrillation’ upgraded the level of recommendation for LAAC to 2a for AF patients with moderate to high risk of stroke and a contraindication to long-term OAC use including spontaneous ICH due to non-reversible cause or serious bleeding related to current falls when cause of falls is not felt to be treatable.35 Our results support the scientific validity of these upgraded recommendations.

    In addition to patients with prior ICH, we had a well-defined patient population who never had an ICH but was still considered at high ICH risk due to the presence of CMBs with or without cSS on brain MRI (more than 88% had three or more MRI markers for high haemorrhagic risk).36 To our knowledge, this is the largest group of patients who underwent LAAC for this indication reported in the literature. Studies performed in post-stroke anticoagulated AFib patient populations as well as community-based studies of older adults show a 4–6-folds increase in ICH risk when CMBs are present.6 37 Annual risk of first time ICH reaches 5 per 100 patient-years among patients with strictly cortical microbleeds that fulfil criteria for CAA without a prior ICH.38 In our cohort, 52.1% had cSVD imaging markers consistent with the diagnosis of CAA according to the Boston criteria V.2.0. None of these CMB-group patients had a spontaneous ICH and only one patient suffered a T-ICH 3 months after LAAC (IR 1.9% per 100 patient-years). The small number of patients in the CMB-group limits our ability to draw firm conclusions, but our results still support the view that LAAC is an important consideration for ischaemic and haemorrhagic stroke prevention in an NVAF population with CMB(s) but no prior ICH. This hypothesis should be tested in RCTs.

    Optimal antithrombotic regimen after LAAC is a debatable topic especially in patients at high ICH risk. The two RCTs, which laid the grounds for the approval of the Watchman device by the FDA, did not include patients who had contraindications to OAC and prescribed OAC for 6 weeks post-LAAC.10 11 In 2022, the FDA published revised labelling to the Watchman-FLX allowing a 45-day dual antiplatelet therapy option as an alternative to warfarin or DOAC use.39 In our study, 44 (30.1%) patients were discharged after LAAC on antiplatelet therapy only and at 1 year only 10 (6.8%) patients were still using OAC. There is a high level of variability in post-LAAC antithrombotic use and available data including our study show acceptable results with individualised risk assessment and treatment among NVAF patients who have different ICH risks.40

    The main limitation of this study is its retrospective, observational design but all of our patients received standardised clinical, and imaging follow-up in the same hospital system, which improved the quality and consistency of longitudinal data acquisition. Additionally, the absence of a comparator group hinders direct comparisons with other treatment approaches, but such a comparator group is hard to identify as patients who did not undergo a procedure do not necessarily get standardised follow-up and they might have lower ischaemic/haemorrhagic risks. To address this limitation, we have used information from both large-scale observational studies and available RCTs as a point of reference. Detailed information on ongoing RCTs for ICH survivors with NVAF has been provided in a recent consensus statement on atrial fibrillation, but these studies proved to be very slow to enrol patients because of very high mortality rate of OAC-related ICHs, so well-conducted observational studies are still important.33

    Conclusion

    Our study results show that high ICH risk NVAF patients who undergo LAAC exhibit low incidence rates of ICH when compared with patients randomised to (D)OACs in similar populations and low rates of AIS based on CHA₂DS₂-VASc. These findings support the view that NVAF patients with a high ICH risk profile are appropriate candidates for LAAC as a preferable option over lifelong OAC treatment for ischaemic and haemorrhagic stroke prevention.

    Data availability statement

    Anonymized data not published within this article will be made available by reasonable request from a qualified investigator.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants but IRB protocol #2021P003370 exempted this study. This was a retrospective study for which IRB waived the need for informed consent.

    References

    1. ↵
      1. Pasi M,
      2. Charidimou A,
      3. Boulouis G, et al
      . Mixed-location cerebral hemorrhage/Microbleeds: underlying microangiopathy and recurrence risk. Neurology 2018;90:e119–26. doi:10.1212/WNL.0000000000004797
      OpenUrlCrossRefPubMed
    2. ↵
      1. Charidimou A,
      2. Boulouis G,
      3. Roongpiboonsopit D, et al
      . Cortical superficial Siderosis Multifocality in cerebral Amyloid Angiopathy: A prospective study. Neurology 2017;89:2128–35. doi:10.1212/WNL.0000000000004665
      OpenUrlPubMed
    3. ↵
      1. Charidimou A,
      2. Karayiannis C,
      3. Song T-J, et al
      . Brain Microbleeds, anticoagulation, and hemorrhage risk: meta-analysis in stroke patients with AF. Neurology 2017;89:2317–26. doi:10.1212/WNL.0000000000004704
      OpenUrl
    4. ↵
      1. Al-Shahi Salman R,
      2. Keerie C,
      3. Stephen J, et al
      . Effects of oral anticoagulation for atrial fibrillation after spontaneous intracranial haemorrhage in the UK: a randomised, open-label, assessor-masked, pilot-phase, non-inferiority trial. The Lancet Neurology 2021;20:842–53. doi:10.1016/S1474-4422(21)00264-7
      OpenUrl
    5. ↵
      1. Schreuder F,
      2. van Nieuwenhuizen KM,
      3. Hofmeijer J, et al
      . Apixaban versus no anticoagulation after anticoagulation-associated intracerebral haemorrhage in patients with atrial fibrillation in the Netherlands (APACHE-AF): a randomised, open-label, phase 2 trial. Lancet Neurol 2021;20:907–16. doi:10.1016/S1474-4422(21)00298-2
      OpenUrlPubMed
    6. ↵
      1. Wilson D,
      2. Ambler G,
      3. Shakeshaft C, et al
      . Cerebral Microbleeds and intracranial haemorrhage risk in patients Anticoagulated for atrial fibrillation after acute ischaemic stroke or transient ischaemic attack (CROMIS-2): a Multicentre observational cohort study. Lancet Neurol 2018;17:539–47. doi:10.1016/S1474-4422(18)30145-5
      OpenUrlCrossRefPubMed
    7. ↵
      1. Martí-Fàbregas J,
      2. Medrano-Martorell S,
      3. Merino E, et al
      . MRI predicts intracranial hemorrhage in patients who receive long-term oral anticoagulation. Neurology 2019;92:e2432–43. doi:10.1212/WNL.0000000000007532
      OpenUrlPubMed
    8. ↵
      1. Gokcal E,
      2. Horn MJ,
      3. Gurol ME
      . The role of biomarkers and neuroimaging in ischemic/hemorrhagic risk assessment for cardiovascular/cerebrovascular disease prevention. Handb Clin Neurol 2021;177:345–57. doi:10.1016/B978-0-12-819814-8.00021-4
      OpenUrl
    9. ↵
      1. Das AS,
      2. Gökçal E,
      3. Regenhardt RW, et al
      . Clinical and neuroimaging risk factors associated with the development of intracerebral hemorrhage while taking direct oral anticoagulants. J Neurol 2022;269:6589–96. doi:10.1007/s00415-022-11333-2
      OpenUrlCrossRef
    10. ↵
      1. Reddy VY,
      2. Sievert H,
      3. Halperin J, et al
      . Percutaneous left atrial appendage closure vs warfarin for atrial fibrillation: a randomized clinical trial. JAMA 2014;312:1988–98. doi:10.1001/jama.2014.15192
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Holmes DR,
      2. Kar S,
      3. Price MJ, et al
      . Prospective randomized evaluation of the watchman left atrial appendage closure device in patients with atrial fibrillation versus long-term warfarin therapy: the PREVAIL trial. J Am Coll Cardiol 2014;64:1–12. doi:10.1016/j.jacc.2014.04.029
      OpenUrlFREE Full Text
    12. ↵
      1. Gurol ME
      . Nonpharmacological management of atrial fibrillation in patients at high intracranial hemorrhage risk. Stroke 2018;49:247–54. doi:10.1161/STROKEAHA.117.017081
      OpenUrlFREE Full Text
    13. ↵
      1. Lakkireddy D,
      2. Thaler D,
      3. Ellis CR, et al
      . Amplatzer amulet left atrial appendage Occluder versus watchman device for stroke prophylaxis (amulet IDE): A randomized, controlled trial. Circulation 2021;144:1543–52. doi:10.1161/CIRCULATIONAHA.121.057063
      OpenUrlCrossRefPubMed
    14. ↵
      1. Das AS,
      2. Regenhardt RW,
      3. Vernooij MW, et al
      . Asymptomatic cerebral small vessel disease: insights from population-based studies. J Stroke 2019;21:121–38. doi:10.5853/jos.2018.03608
      OpenUrlCrossRefPubMed
    15. ↵
      1. Wardlaw JM,
      2. Smith EE,
      3. Biessels GJ, et al
      . Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration. Lancet Neurol 2013;12:822–38. doi:10.1016/S1474-4422(13)70124-8
      OpenUrlCrossRefPubMedWeb of Science
    16. ↵
      1. Greenberg SM,
      2. Vernooij MW,
      3. Cordonnier C, et al
      . Cerebral Microbleeds: a guide to detection and interpretation. Lancet Neurol 2009;8:165–74. doi:10.1016/S1474-4422(09)70013-4
      OpenUrlCrossRefPubMedWeb of Science
    17. ↵
      1. Charidimou A,
      2. Linn J,
      3. Vernooij MW, et al
      . Cortical superficial Siderosis: detection and clinical significance in cerebral Amyloid Angiopathy and related conditions. Brain 2015;138:2126–39. doi:10.1093/brain/awv162
      OpenUrlCrossRefPubMed
    18. ↵
      1. Fazekas F,
      2. Kleinert R,
      3. Offenbacher H, et al
      . Pathologic correlates of incidental MRI white matter signal Hyperintensities. Neurology 1993;43:1683–9. doi:10.1212/wnl.43.9.1683
      OpenUrlCrossRefPubMed
    19. ↵
      1. Charidimou A,
      2. Boulouis G,
      3. Haley K, et al
      . White matter Hyperintensity patterns in cerebral Amyloid Angiopathy and hypertensive Arteriopathy. Neurology 2016;86:505–11. doi:10.1212/WNL.0000000000002362
      OpenUrlCrossRefPubMed
    20. ↵
      1. Gokcal E,
      2. Horn MJ,
      3. van Veluw SJ, et al
      . Lacunes, Microinfarcts, and vascular dysfunction in cerebral Amyloid Angiopathy. Neurology 2021;96:e1646–54. doi:10.1212/WNL.0000000000011631
      OpenUrlPubMed
    21. ↵
      1. Charidimou A,
      2. Boulouis G,
      3. Pasi M, et al
      . MRI-visible Perivascular spaces in cerebral Amyloid Angiopathy and hypertensive Arteriopathy. Neurology 2017;88:1157–64. doi:10.1212/WNL.0000000000003746
      OpenUrlCrossRefPubMed
    22. ↵
      1. Best JG,
      2. Barbato C,
      3. Ambler G, et al
      . Association of enlarged Perivascular spaces and anticoagulant-related intracranial hemorrhage. Neurology 2020;95:e2192–9. doi:10.1212/WNL.0000000000010788
      OpenUrl
    23. ↵
      1. Charidimou A,
      2. Boulouis G,
      3. Frosch MP, et al
      . The Boston criteria version 2.0 for cerebral Amyloid Angiopathy: a Multicentre, retrospective, MRI-neuropathology diagnostic accuracy study. Lancet Neurol 2022;21:714–25. doi:10.1016/S1474-4422(22)00208-3
      OpenUrlCrossRefPubMed
    24. ↵
      1. Das AS,
      2. Gokcal E,
      3. Biffi A, et al
      . Mechanistic implications of cortical superficial Siderosis in patients with mixed location intracerebral hemorrhage and cerebral Microbleeds. Neurology 2023;101:e636–44. doi:10.1212/WNL.0000000000207476
      OpenUrlCrossRef
    25. ↵
      1. Tsai H-H,
      2. Pasi M,
      3. Tsai L-K, et al
      . Microangiopathy underlying mixed-location intracerebral hemorrhages/Microbleeds: A Pib-PET study. Neurology 2019;92:e774–81. doi:10.1212/WNL.0000000000006953
      OpenUrlPubMed
    26. ↵
      1. Asch CJ,
      2. Luitse MJ,
      3. Rinkel GJ, et al
      . Incidence, case fatality, and functional outcome of intracerebral haemorrhage over time, according to age, sex, and ethnic origin: a systematic review and meta-analysis. Lancet Neurol 2010;9:167–76.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Shoamanesh A,
      2. Committee E-A
      . Anticoagulation in patients with cerebral Amyloid Angiopathy. Lancet 2023;402:1418–9. doi:10.1016/S0140-6736(23)02025-1
      OpenUrl
    28. ↵
      1. Gilhofer TS,
      2. Nestelberger T,
      3. Kang M, et al
      . Stroke prevention with left atrial appendage closure in patients with atrial fibrillation and prior intracranial hemorrhage. CJC Open 2023;5:404–11. doi:10.1016/j.cjco.2023.03.004
      OpenUrl
    29. ↵
      1. Boersma LV,
      2. Ince H,
      3. 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 Arrhythm Electrophysiol 2019;12:e006841. doi:10.1161/CIRCEP.118.006841
    30. ↵
      1. Ajmal M,
      2. Sipra Q,
      3. Pecci C, et al
      . Feasibility of left atrial appendage closure in atrial fibrillation patients with a history of intracranial bleeding: A systematic review of observational studies. J Interv Cardiol 2020;1575839.
    31. ↵
      1. Lehtola H,
      2. Palomäki A,
      3. Mustonen P, et al
      . Traumatic and spontaneous intracranial hemorrhage in atrial fibrillation patients on warfarin. Neurol Clin Pract 2018;8:311–7. doi:10.1212/CPJ.0000000000000491
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Friberg L,
      2. Rosenqvist M,
      3. Lip GYH
      . Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation: the Swedish atrial fibrillation cohort study. Eur Heart J 2012;33:1500–10. doi:10.1093/eurheartj/ehr488
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Gurol ME,
      2. Wright CB,
      3. Janis S, et al
      . Stroke prevention in atrial fibrillation: our current failures and required research. Stroke 2024;55:214–25. doi:10.1161/STROKEAHA.123.040447
      OpenUrl
    34. ↵
      1. Saliba WI,
      2. Kawai K,
      3. Sato Y, et al
      . Enhanced Thromboresistance and Endothelialization of a novel Fluoropolymer-coated left atrial appendage closure device. JACC Clin Electrophysiol 2023;9:1555–67. doi:10.1016/j.jacep.2023.04.013
      OpenUrl
    35. ↵
      1. Joglar JA,
      2. Chung MK,
      3. Armbruster AL, et al
      . ACC/AHA/ACCP/HRS guideline for the diagnosis and management of atrial fibrillation: A report of the American college of cardiology/American heart Association joint committee on clinical practice guidelines. Circulation 2024;149:e1–156. doi:10.1161/CIR.0000000000001193
      OpenUrlCrossRef
    36. ↵
      1. Charidimou A,
      2. Boulouis G,
      3. Xiong L, et al
      . Cortical superficial Siderosis and first-ever cerebral hemorrhage in cerebral Amyloid Angiopathy. Neurology 2017;88:1607–14. doi:10.1212/WNL.0000000000003866
      OpenUrl
    37. ↵
      1. Akoudad S,
      2. Portegies MLP,
      3. Koudstaal PJ, et al
      . Cerebral Microbleeds are associated with an increased risk of stroke: the Rotterdam study. Circulation 2015;132:509–16. doi:10.1161/CIRCULATIONAHA.115.016261
      OpenUrlAbstract/FREE Full Text
    38. ↵
      1. van Etten ES,
      2. Auriel E,
      3. Haley KE, et al
      . Incidence of symptomatic hemorrhage in patients with Lobar Microbleeds. Stroke 2014;45:2280–5. doi:10.1161/STROKEAHA.114.005151
      OpenUrlAbstract/FREE Full Text
    39. ↵
      1. Freeman JV,
      2. Higgins AY,
      3. Wang Y, et al
      . Antithrombotic therapy after left atrial appendage occlusion in patients with atrial fibrillation. J Am Coll Cardiol 2022;79:1785–98. doi:10.1016/j.jacc.2022.02.047
      OpenUrlCrossRefPubMed
    40. ↵
      1. Hutt E,
      2. Wazni OM,
      3. Saliba WI, et al
      . Left atrial appendage closure device implantation in patients with prior intracranial hemorrhage. Heart Rhythm 2019;16:663–8. doi:10.1016/j.hrthm.2018.11.022
      OpenUrl

    Supplementary material

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • X @alvindasMD, @guroledip

    • Contributors All authors fulfilled the criteria for authorship.The guarantor of this study was AAF and takesfull responsibility for the finished work and conduct of the study.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests Dr. Gurol received research grants from National Institute of Health (NIH, RO1NS11452, NS083711).

      Dr. Gurol’s hospital received research grants from AVID, Pfizer and Boston Scientific Corporation.

    • Provenance and peer review Not commissioned; externally peer-reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.