IJC Heart & Vasculature 38 (2022) 100941 Contents lists available at ScienceDirect IJC Heart & Vasculature journal homepage: www.sciencedirect.com/journal/ijc-heart-and-vasculature Impact of operatoŕs experience on peri-procedural outcomes with Watchman FLX: Insights from the FLX-SPA registry2 Ignacio Cruz-González a, b, 1, Francisco Torres Saura c, 1, Blanca Trejo-Velasco a, *, José Antonio Fernández Díaz d, Ricardo Fajardo Molina e, Raquel del Valle-Fernández f, Gerardo Moreno Terribas g, h, David Martí Sánchez i, j, José-Ramón López-Mínguez k, l, Ivan Gomez-Blazquez m, Juan-Carlos Sanmartin Pena n, Javier Botas o, p, Pedro Martin Lorenzo q, Jorge Palazuelos r, Ramón Albarrán Rincon s, Mohsen Mohandes t, Felipe-Jose Rodriguez Entem u, Gerard Martí v, Ernesto Valero w, Hipólito Gutiérrez b, x, Ignacio J. Amat-Santos b, x, Luis Nombela-Franco y, Pablo Salinas y, Luis Teruel z, Joan-Antoni Gomez-Hospital z, Dabit Arzamendi aa, Mario Torres Sanabria aa, Germán Calle Pérez ab, Dolores Cañadas Pruaño ab, Armando Pérez de Prado ac, Tomás Benito González ac, Eduardo Arroyo-Úcar c, Rodrigo Estévez-Loureiro ad, Berenice Caneiro-Queija ad, Jose L. Ibañez Criado ae, Juan M. Ruiz-Nodar ae, af a Cardiology Department, University Hospital of Salamanca, Institute of Biomedical Research of Salamanca, Spain b Biomedical Research Networking Center on Cardiovascular Diseases (CIBERCV), Madrid, Spain c Interventional Cardiology Unit, University Hospital del Vinalopó, Elche, Spain d Interventional Cardiology Unit, University Hospital Puerta de Hierro, Madrid, Spain e Cardiology Department, University Hospital Torrecárdenas, Almería, Spain f Cardiology Department, University Hospital Central de Asturias, Oviedo, Spain g Interventional Cardiology Unit, Cardiology Department, University Hospital Clinico San Cecilio Granada, Spain h Medical School, University of Granada, Spain i Cardiology Department, Hospital Central de la Defensa, Madrid, Spain j Medical School, University of Alcalá, Madrid, Spain k Cardiology Department, Infanta Cristina University Hospital of Badajoz, Spain l Medical School, University of Extremadura, Badajoz, Spain m Cardiology Department, University Hospital 12 de Octubre, Madrid, Spain n Cardiology Department, Clinic University Hospital of Santiago, Santiago de Compostela, Spain o Cardiology Department, University Hospital Fundación Alcorcón, Madrid, Spain p Rey Juan Carlos University School of Medicine, Madrid, Spain q Interventional Cardiology Unit, Cardiology Department, University Hospital of Gran Canaria Dr. Negrin, Las Palmas de Gran Canaria, Spain r Cardiology Department, Hospital La Luz, Madrid, Spain s Cardiology and Cardiac Surgery Department, University Clinic of Navarra, Pamplona, Spain t Interventional Cardiology Unit, Joan XXIII University Hospital, Tarragona, Spain u Cardiology Department, University Hospital Marqués de Valdecilla, Santander, Spain v Cardiology Department, University Hospital Vall d́ Hebron, Barcelona, Spain w Cardiology Department, Clinic University Hospital of Valencia, Valencia, Spain x Cardiology Department, Clinic University Hospital of Valladolid, Spain y Cardiovascular Institute, Hospital Clínico San Carlos, Madrid, Spain z Cardiology Department, Bellvitge University Hospital-IDIBELL, L’Hospitalet de Llobregat, Barcelona, Spain aa Cardiology Department, University Hospital de la Santa Creu i Sant Pau, Barcelona, Spain ab Cardiology Department, University Hospital Puerta del Mar, Cádiz, Spain ac Cardiology Department, University Hospital of León, Leon, Spain ad Interventional Cardiology Unit, University Hospital Álvaro Cunqueiro, Vigo, Spain Abbreviations: ICE, intracardiac echocardiography; LAA, left atrial appendage; LAAO, left atrial appendage occlusion; OAC, oral anticoagulation; TEE, trans­ esophageal echocardiography. 2 Every author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. * Corresponding author at: San Vicente 182, 37007, Salamanca, Spain, Cardiology Department, Hospital Universitario de Salamanca, Salamanca, Spain. E-mail addresses: treejooblanca@hotmail.com, btv2211@gmail.com (B. Trejo-Velasco). https://doi.org/10.1016/j.ijcha.2021.100941 Received 9 November 2021; Received in revised form 9 December 2021; Accepted 23 December 2021 Available online 29 December 2021 2352-9067/© 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). I. Cruz-González et al. IJC Heart & Vasculature 38 (2022) 100941 ae Cardiology Department, Hospital General Universitario de Alicante, Instituto de Investigación Sanitaria y Biomédica de Alicante, Spain af Facultad de Medicina, Universidad Miguel Hernández, Sant Joan, Alicante, Spain A R T I C L E I N F O A B S T R A C T Keywords: Background: The Watchman FLX is a device upgrade of the Watchman 2.5 that incorporates several design en­ Atrial fibrillation hancements intended to simplify left atrial appendage occlusion (LAAO) and improve procedural outcomes. This Left atrial appendage occlusion study compares peri-procedural results of LAAO with Watchman FLX (Boston Scientific, Marlborough, Massa­ Devices chusetts) in centers with varying degrees of experience with the Watchman 2.5 and Watchman FLX. Outcomes Methods: Prospective, multicenter, “real-world” registry including consecutive patients undergoing LAAO with the Watchman FLX at 26 Spanish sites (FLX-SPA registry). Implanting centers were classified according to the center’s prior experience with the Watchman 2.5. A further division of centers according to whether or not they had performed ≤ 10 or > 10Watchman FLX implants was prespecified at the beginning of the study. Procedural outcomes of institutions stratified according to their experience with the Watchman 2.5 and FLX devices were compared. Results: 359 patients [mean age 75.5 (SD8.1), CHA2DS2-VASc 4.4 (SD1.4), HAS-BLED 3.8(SD0.9)] were included. Global success rate was 98.6%, successful LAAO with the first selected device size was achieved in 95.5% patients and the device was implanted at first attempt in 78.6% cases. There were only 9(2.5%) major peri-procedural complications. No differences in efficacy or safety results according to the centeŕs previous experience with Watchman 2.5 and procedural volume with Watchman FLX existed. Conclusions: The Watchman FLX attains high procedural success rates with complete LAA sealing in unselected, real-world patients, along with a low incidence of peri-procedural complications, regardless of operatoŕs expe­ rience with its previous device iteration or the number of Watchman FLX devices implanted. 1. Introduction 2. Methods Left atrial appendage occlusion (LAAO) is an established non- 2.1. Study population pharmacological alternative to oral anticoagulation (OAC) for stroke prevention in patients with non-valvular atrial fibrillation (AF) at high The FLX-SPA is a multicenter, prospective, open-label registry that bleeding risk [1]. Since conception of this technique, several LAAO included consecutive patients with non-valvular AF undergoing LAAO dedicated devices with differential design features have been developed with the Watchman FLX device (Boston Scientific, Marlborough, Mas­ [1,2]. Amongst them, the Watchman 2.5 device (Boston Scientific, sachusetts, US) in 26 Spanish centers between October 2019 and Marlborough, Massachusetts, US) gathers the greatest body of evidence January 2021. supporting its safety and efficacy, both in the form of randomized clin­ The device was implanted on an all-comer basis in unselected pa­ ical trials and multicenter registries [3–4]. Notwithstanding, LAAO with tients undergoing LAAO. All patients underwent standard diagnostic the Watchman 2.5 device can be challenging in certain left atrial evaluation including estimation of thromboembolic and bleeding risk appendage (LAA) anatomies such as broad and shallow LAA, those with based on CHA2DS2-VASc and HAS-BLED scores. Baseline clinical and a complex internal architecture as well as chicken-wing–shaped LAA imaging characteristics as well as procedural details were collected with a short proximal segment and an acute bend [1,5]. prospectively and registered in a dedicated database at each of the The Watchman FLX is a device upgrade of the previous Watchman participating centers. 2.5 that incorporates substantial design changes aimed to simplify the The study protocol complied with the Helsinki Declaration and was LAAO procedure in a broader range of LAA anatomies. Since its approval approved by the Ethics Committee and all subjects provided informed by CE mark authorities in 2019, several registries have reported prom­ consent prior to the procedure. ising peri-procedural outcomes with the Watchman FLX, including lower need for device recapture and repositioning, reduced number of devices used per procedure and a high degree of complete LAA sealing [6–7]. 2.2. Implanting centers More recently, the PINNACLE FLX trial confirmed consistent favorable results at 1-year follow-up [8]. Notwithstanding, these studies were Specific training for LAAO with the Watchman FLX was conducted performed in centers with a high volume of LAAO procedures, by op­ prior to patientś enrollment in every center. All procedures were per­ erators with prior experience with the Watchman 2.5 device, which formed by interventional cardiologists with at least 2 years prior expe­ could have exerted a positive bias on the results. In order for new centers rience in interventional cardiology. In each center, there were 2 to incorporate the Watchman FLX device into their therapeutical operators involved in LAAO procedures, with the exception of in­ arsenal, it remains relevant to assess if operators without prior experi­ stitutions contributing with ≤ 6 cases, in which only one operator was ence with the Watchman 2.5 device, as well as those with a shorter involved. There was no run-in period and all patients from the first case experience with the Watchman FLX, may achieve similarly favorable performed with the Watchman FLX were included. Implanting centers peri-procedural outcomes [9]. were classified according to their previous experience with the The aim of the multicenter FLX-SPA registry is to report and compare Watchman 2.5 device: centers without prior experience were catego­ peri-procedural outcomes with the Watchman FLX device in centers with rized as “Watchman 2.5 naïve centers”, while institutions that already and without prior experience with the Watchman 2.5 device, as well as had practice with the Watchman 2.5 device were considered separately. in centers with varying degrees of procedural volume with the In addition, implanting centers were assorted in two further groups, Watchman FLX. according to whether or not they had implanted ≤ 10 (initial Watchman FLX experience) or > 10 Watchman FLX devices (established Watchman FLX experience). The cut-off value for comparisons was established at 10 1 Both Authors have contributed to this work in the same amount and should Watchman FLX implants, as this was the minimum number of proced­ both be considered as first authors. ures deemed necessary to become proficient with this device. Baseline 2 I. Cruz-González et al. IJC Heart & Vasculature 38 (2022) 100941 Fig. 1. Patientś distribution and procedural outcomes according to centers experience and procedural volume. A. 320(81%) patients underwent LAAO at centers with prior experience with the Watchman 2.5 device, while the remainder 39(19%) were treated at “Watchman 2.5 naïve” centers. B. 267(74%) patients were managed at institutions that had performed over 10 Watchman FLX implants, whereas 92(26 %) were treated at institutions that had performed ≤ 10 procedures. Device recapture was more frequent in centers that had performed ≤ 10 Watchman FLX implants. No other differences in procedural details between both groups of institutions existed. *P-value < 0.001. P-value > 0.05 for all other comparisons. clinical and imaging features, procedural details and peri-procedural of the circumflex coronary artery, to a point 1–2 cm distal to the left results of both groups were compared. upper pulmonary vein ridge. Patients with LAA thrombus, prior LAA surgical ligation as well as those with a LAA anatomy unsuitable for Watchman FLX implantation as per manufactureŕs instructions of use 2.3. Watchman FLX device (LZ diameters < 17 mm or > 31 mm and LAAs shorter than the diameter of the required occlusion device) were excluded. No other morpholog­ Specific design features of the Watchman FLX device have been ical exclusion criteria for Watchman FLX implantation existed. Device previously described [6–8]. Main refinements in the design of the sizing was selected according to the maximal LZ diameter, with over­ Watchman FLX in comparison to the Watchman 2.5 include a 10–20% sizing aimed to attain final device compression between 10 and 30%. length reduction and a closed end configuration that facilitate device implantation in shallow LAA (minimum depth required is 50% of device size). Additionally, device advancement within the LAA can be per­ 2.5. Procedural details and outcomes formed while partially deployed in the “ball” configuration, thanks to its atraumatic distal end with a fluoroscopic marker that reduces the risk of LAAO was performed under general anesthesia or conscious seda­ distal perforations. Furthermore, polyester fabric coverage has been tion, employing transesophageal echocardiography (TEE), micro-TEE or extended to reduce peri-device leaks, the delivery cable screw recessed intracardiac echocardiography (ICE) according to local practice. Tech­ to reduce device related thrombus (DRT) and fixation mechanisms nical success was defined as complete LAA exclusion with no peri-device enhanced, with the addition of an extra row of 18 J-shaped anchors. leak > 5 mm on color-Doppler TEE and no device-related complications. Peri-procedural outcomes were recorded during the first 7 days after 2.4. Pre-procedural imaging LAAO. Major peri-procedural adverse events were registered according to the definitions in the Munich Consensus Document on LAAO [10] and Pre-procedural evaluation of the LAA was performed with TEE in all included, but were not limited to, device embolization and thrombosis, cases and its morphology and dimensions registered. The landing zone cardiac perforation, tamponade and pericardial effusion requiring (LZ) was measured from the inferior part of the ostial plane at the level intervention, major bleeding defined as type ≥ 3 of Bleeding Academic 3 I. Cruz-González et al. IJC Heart & Vasculature 38 (2022) 100941 Table 1 Supplementary Tables 1 and 2. Baseline clinical characteristics. Total sample 3.1. Peri-procedural details (n ¼ 359) Age(years) 75.5 ± 8.1 Imaging and procedural characteristics according to centerś experi­ Sex, male(n, %) 218 (60.7) ence and procedural volume are summarized in Tables 2 and 3. BMI,kg/m2* 27.9 ± 4.5 Mean maximal and minimal LZ diameters were 20.9 ± 3.8 mm and Arterial hypertension*(n, %) 272(86.9) 18.1 (SD 3.5) mm, respectively, range 12 to 34 mm, and mean LAÁ s Diabetes mellitus*(n, %) 107(34.2) Coronary artery disease*(n, %) 77(24.6) depth was 25.1 (SD 6.1) mm. In 54 (15.1%) of patients, the LAÁ s depth Heart failure*(n, %) 92(29.4) was shorter than the maximal LZ diameter, a factor that would have LVEF,% 57.5 ± 9.6 precluded LAAO with the Watchman 2.5 device. In addition, the LZ Chronic kidney disease(n, %) 91(25.3) diameter exceeded 31 mm in 2 (0.6%) patients and was ≤ 17 mm in 44 Dialysis(n, %) 34(9.5) (12.3%) of cases. Overall, 78 (21.7%) of patients undergoing LAAO with Arterial vascular disease*(n, %) 69(22.0) Prior ischemic stroke/transient ischemic attack*(n, %) 133(42.5) the Watchman FLX in our sample would have been unsuitable candi­ Intracranial hemorrhage(n, %) 100(27.9) dates for LAAO with the Watchman 2.5, according to the instructions for Prior bleeding(n, %) 290(80.8) use. Recurrent bleeding(n, %) 159(44.4) LAAO was performed under general anhaesthesia in 201 (56%) pa­ Contraindication to OAC(n, %) Absolute 147(40.9) tients. The majority of procedures were guided by TEE (68.5%), while Relative 163(45.4) micro-TEE and ICE were used in 27.6% and 3.4% of patients, respec­ None 49(13.6) tively. Centers that had performed ≤ 10 Watchman FLX implants Indication for LAAO†(n, %) employed general anesthesia more frequently (83.7% versus 46.4% p = Bleeding 290(80.8) 0.001), while use of micro-TEE was more common in centers that had High bleeding risk 38(10.6) Stroke despite OAC 41(11.4) performed > 10 Watchman FLX implants (36% versus 3.3%, p = 0.001) Labile INR 46(12.8) as well as in those with prior experience with the Watchman 2.5 device High risk falls 6(1.7) (30.3% versus 5.1, p = 0.001). Overall, 7 (1.9%) patients underwent CHA2DS2 –VASc Score 4.4 ± 1.4 combined procedures with LAAO (1 transcatheter aortic valve implan­ HAS-BLED Score 3.8 ± 0.9 tation, 2 percutaneous edge-to-edge mitral valve repair and 4 pulmo­ Values: mean ± SD or n(%). nary vein ablation), without differences according to the hospitalś *N = 313. †: All indications that applied in every single patient were recorded. characteristics. BMI:body mass index;GI: gastrointestinal; LAAO:left atrial appendage occlusion; The most frequently implanted device size was Watchman FLX 27 LVEF:left ventricular ejection fraction; OAC:oral anticoagulation. mm, in 38.2% of patients. Technical success was achieved in 354 (98.6%) cases, without differences between centers, Fig. 1. Successful Research Consortium (BARC), major vascular complications, stroke or LAAO with the first selected device size was achieved in 343 (95.5%) transient ischemic attack, systemic embolism, myocardial infarction and cases and the device was implanted at first attempt in 282 (78.6%) pa­ death. Peri-device leaks were classified according to the width of the tients. Device recapture and repositioning was significantly more color-Doppler jet on TEE as follows: mild (1–3 mm), moderate (>3mm frequent in centers that had performed ≤ 10 Watchman FLX implants, but ≤ 5 mm) and severe (>5mm) [10]. Antithrombotic treatment at whilst “Watchman 2.5 naïve centers” employed a greater volume of discharge was adjusted according to individual bleeding and ischemic contrast than centers with prior experience. risk, at each of the participating centers. 3.2. Peri-procedural and in-hospital outcomes 2.6. Statistical analysis Major peri-procedural and in-hospital complications recorded during Categorical variables were expressed as frequencies and percentages, the first 7-days post-procedure occurred in 9 (2.5%) patients, as depicted and continuous variables as mean (SD) or median (interquartile range). in Tables 4 and 5, Fig. 2. No patient died within index hospitalization. Comparisons between centres were performed by means of Chi-square BARC ≥ 3 bleedings were the most frequent complication in 4 (1.1%) tests or Fisher exact tests for categorical variables and T-Student and patients and there was just 1 (0.3%) pericardial tamponade requiring Mann–Whitney U-tests for continuous variables, as appropriate, after pericardiocentesis. Acute device-related complications included 1 testing for normality by means of the Shapiro-Wilk test. Statistical an­ (0.3%) early DRT that was successfully managed with low-molecular alyses were performed with SPSS 21.0 for Windows (IBM, Chicago, weight heparin for 4 weeks. There were no cases of device embolisa­ Illinois). tion. Complete LAA sealing post-procedure was confirmed by TEE in 344 (97.2%) patients and there were no severe (>5mm) peri-device leaks. No 3. Results differences in peri-procedural complications between centers existed. Mean length of hospital stay was 1.8 (SD 1.8) days. At discharge, 164 359 consecutive patients from 26 participating centers were (45.7%) were managed with dual antiplatelet therapy, 73 (20.3%) with included. The distribution of patients according to the centerś experi­ single antiplatelet therapy, 86 (23.9%) with anticoagulation alone and ence with the Watchman 2.5 device and the centeŕs procedural volume 27 (7.5%) with an antiplatelet agent on top of anticoagulation, Table 4 with Watchman FLX is depicted in the Fig. 1 and Supplementary and 5. Figure 1. The main baseline clinical characteristics of included patients are 4. Discussion displayed in Table 1. Mean age was 75.5 (SD 8.1) years and mean CHA2DS2-VASc and HAS-BLED Scores were 4.4 (SD 1.4) and 3.8 (SD Our study provides further evidence on procedural safety and effi­ 0.9), respectively. Overall, 133 (42.%) of patients had a history of prior cacy of LAAO with the Watchman FLX device in real-world practice with ischemic stroke or transient ischemic attack, while previous bleeding similarly favorable results regardless of the centerś degree of expertise. complications were present in 289 (80.5%) patients. Baseline features of Its main findings can be summarized as follows: first, technical patients according to the centers’ prior experience with the Watchman success rates with the Watchman FLX were remarkably high at 98.6%, 2.5 device and volume of Watchman FLX implants are depicted in regardless of LAA anatomy and there were no cases of severe (>5 mm) 4 I. Cruz-González et al. IJC Heart & Vasculature 38 (2022) 100941 Table 2 Peri-procedural characteristics according to centerś experience with Watchman 2.5 device. No prior experience Watchman Prior experience Watchman Total Difference of the means (95% p- 2.5 2.5 sample CI) value (n ¼ 39) (n ¼ 320) (n ¼ 359) LAA morphology(n,%)* 0.030 Chicken-wing 9(37.5) 67(27.3) 76(28.4) Windsock 9(37.5) 116(47.3) 124(46.3) Cauliflower 1(4.2) 45(18.4) 46(17.2) Cactus 5(20.8) 17(6.9) 22(8.2) LAA landing zone(mm) Maximal diameter 20.6 ± 4.5 20.9 ± 3.8 20.9 ± 3.8 − 0.26 (-1.9, +1.3) 0.946 Minimal diameter 17.2 ± 3.8 18.2 ± 3.5 18.1 ± 3.5 − 0.98 (-2.3, +0.4) 0.215 LAA length(mm) 27.3 ± 8.5 24.9 ± 5.7 25.1 ± 6.1 0.341 Unsuitability for Watchman 2.5 device 9(23.1) 69(21.6) 78 (21.7) 0.829 (n,%) Procedural anhaesthesia(n,%) 0.182 Superficial sedation 14(35.9) 144(45) 158(44) General anhaesthesia 25(64.1) 176(55) 201(56) Procedural imaging guidance(n,%) 0.001 TEE Micro-TEE 36(92.3) 210(65.6) 246(68.5) ICE 2(5.1) 97(30.3) 99(27.6) 1(2.6) 13(4.1) 14(3.9) Combined procedure(n,%) 1(2.6) 6(1.9) 7(1.9) 0.556 Size of implanted device 0.689 FLX-20mm 7(17.9) 35(10.9) 42(11.7) FLX-24mm 13(33.3) 124(38.8) 137(38.2) FLX-27mm 12(30.8) 89(27.8) 101(28.1) FLX-31mm 6(15.4) 56(17.5) 59(16.4) FLX-35 mm 1(2.6) 16(5) 17(4.8) Technical success(n,%) 39(100) 315(98.4) 354(98.6) 0.561 >1 device per patient(n,%) 2(5.1) 13(4.1) 15(4.2) 0.754 Devices per patient 1.0 ± 0.2 1.1 ± 0.2 1.0 ± 0.2 0.007(-0.06,+0.08) 0.903 Device recapture(n,%) 7(17.9) 70(21.9) 77(21.4) 0.682 Device recaptures/patient 0.8 ± 1.3 1.0 ± 1.1 0.9 ± 1.1 − 0.2 (-0.8, +0.3) 0.182 Device compression(%) 23.1 ± 12.3 27.2 ± 12.4 26.8 ± 12.4 − 4.1(-9.2, +1.0) 0.118 Fluoroscopy time(minutes) 12(6–18) 14(9–20) 14(9–20) − 0.4(-5.1, +4.3) 0.142 Contrast media volume(ml) 112(85–134) 75.5(48.8–105) 80(52–107) 34.9 (7.1,62.7) <0.001 Values:mean ± SD, median (interquartile range) or n(%).*N = 269. CI: Confidence Interval. ICE:intra-cardiac echocardiography; LAA:left atrial appendage; TEE:transesophageal echocardiography. peri-device leaks on post-procedural TEE. Second, LAAO with the can be simplified with the Watchman FLX, as compared to its prior de­ Watchman FLX was safe, with a low incidence of major peri-procedural vice iteration. adverse events during the first 7 days after LAAO (2.5%) and no cases of Moreover, there were no severe peri-device leaks and mild-to- device embolization or death. Third, operators with no experience with moderate leaks occurred in only 2.8% of patients. Although our study the previous device iteration, the Watchman 2.5, as well as those that does not provide follow-up TEE data, lack of progression from mild to had performed ≤ 10 Watchman FLX implants, attained similar technical severe leaks described in prior reports along with minor variations in the success and peri-procedural complication rates than more experienced prevalence of peri-device leaks during follow-up after Watchman FLX operators. implantation are comforting and suggest that post-procedural outcomes The Watchman FLX has recently emerged as a valuable option for observed in the current study are likely to maintain over time [6,8]. LAAO that enables a simpler implantation procedure in a wider span of Regarding peri-procedural safety, we observed 9 (2.5%) major pro­ LAA anatomies than its previous device iteration, thanks to several cedural and in-hospital complications, an outcome largely determined design enhancements [6–7]. These finding were recently confirmed in by post-procedural bleeding events, which occurred in 4 (1.1%) pa­ the PINNACLE FLX randomized trial [8]. tients. Our safety results are similar to those reported in previous reg­ Our registry provides similarly high technical success rates as pre­ istries with the Watchman FLX but substantially higher than those vious studies in a large sample of real-world patients with unselected described in the PINNACLE FLX study [8]. However, the latter did not LAA anatomies, including 21.7% of cases that would have been include bleeding events into the primary safety endpoint and targeted perceived as technically challenging or unsuitable for LAAO with the patients with a substantially lower bleeding risk, as predicted by HAS- Watchman 2.5. Of note, complete LAA sealing was achieved with the BLED Score [3.8 (SD 0.9) in the SPA-FLX registry versus 2.0 (SD 1.0) first selected device size in 95.5% of cases, with a low number of device in the PINNACLE FLX trial], two factors likely to have influenced these recaptures per patient at 0.9 (SD 1.1). This value is lower than previ­ findings. Of importance, there was only one pericardial effusion in our ously described for the Watchman 2.5 [11–12], despite a substantial study which was successfully managed with percutaneous drainage and proportion of cases being performed at centers with limited experience no cases of death or device embolization, a complication that led to with the device. Altogether, these results support that LAAO procedure market retrieval of a previous version of the Watchman FLX in 2017, but 5 I. Cruz-González et al. IJC Heart & Vasculature 38 (2022) 100941 Table 3 Peri-procedural characteristics according to the centerś procedural volume. Centers with ≤ 10 Watchman FLX Centers with > 10 Watchman FLX Total Difference of the means p- implants implants sample (95% CI) value (n ¼ 92) (n ¼ 267) (n ¼ 359) LAA morphology(n,%)* 0.039 Chicken-wing 23(31.9) 53(26.9) 76(28.4) Windsock 28(38.9) 97(49.2) 124(46.3) Cauliflower 10(13.9) 36(18.3) 46(17.2) Cactus 11(15.3) 11(5.6) 22(8.2) LAA landing zone(mm) Maximal diameter 20.6 ± 3.9 21 ± 3.8 20.9 ± 3.8 − 0.4(-1.5, +0.60) 0.401 Minimal diameter 17.9 ± 3.3 18.2 ± 3.6 18.1 ± 3.5 − 0.3(-1.2, +0.60) 0.500 LAA length(mm) 26.4 ± 6.2 24.5 ± 5.9 25.1 ± 6.1 0.021 Unsuitability for Watchman 2.5 22(23.9) 56(21) 78(21.7) 0.555 device(n,%) Procedural anhaesthesia (n,%) <0.001 Superficial sedation 15(16.3) 143(53.6) 158(44) General anhaesthesia 77(83.7) 124(46.4) 201(56) Procedural imaging guidance(n,%) <0.001 TEE Micro-TEE 87(94.6) 159(59.6) 246(68.5) ICE 3(3.3) 96(36) 99(27.6) 2(2.2) 12(4.5) 14(3.9) Combined procedure(n,%) 4(4.3) 3(1.1) 7(1.9) 0.074 Size of implanted device 0.537 FLX-20mm 12(13) 30(11.2) 42(11.7) FLX-24mm 36(39.1) 101(37.8) 137(38.2) FLX-27mm 29(31.5) 72(27) 101(28.1) FLX-31mm 13(14.1) 49(18.4) 59(16.4) FLX-35 mm 2(2.2) 15(5.6) 17(4.8) Technical success(n,%) 90(97.8) 264(98.9) 354(98.6) 0.606 >1 device per patient(n,%) 5(5.4) 10(3.7) 15(4.2) 0.485 Devices per patient 1.1 ± 0.2 1.0 ± 0.2 1.0 ± 0.2 0.01 (-0.04,+0.06) 0.351 Device recapture(n,%) 34(37) 43(16.1) 77(21.4) <0.001 Device recaptures/patient 1 ± 1.2 0.9 ± 1.0 0.9 ± 1.1 0.07 (-0.3, +0.45) 0.718 Device compression(%) 25.7 ± 12.7 27.3 ± 12.3 26.8 ± 12.4 − 1.6 (-4.9, +1.7) 0.354 Fluoroscopy time(minutes) 16(8–23) 14(9–19) 14(9–20) 2.8(-0.6, +6.2) 0.281 Contrast-media volume(ml) 80(50–100) 80(52.8–119) 80(52–107) − 14.3 (-27.7, − 0.83) 0.087 Values:mean ± SD, median (interquartile range) or n(%).*N = 269. CI: Confidence Interval. ICE:intra-cardiac echocardiography; LAA:left atrial appendage; TEE:transesophageal echocardiography. which has not been reported in studies with the newer generation of the including its ability to be totally recaptured and repositioned both device [6–8]. proximally as well as distally in the “ball” configuration and the possi­ Our study provides an important contribution to existing evidence as bility to perform rotational adjustments during deployment, without the it supports the adoption of the Watchman FLX by less experienced need to completely retrieve or change the device. centers, with similar procedural safety and efficacy outcomes than more In summary, the results of the present study reaffirm the outstanding expert institutions. This finding is relevant given the continuous raise in device performance of the Watchman FLX in unselected, real-world LAAO procedures along with LAAO operators and implanting centres patients. In addition, the Watchman FLX emerges as an attractive op­ that is taking place across the world in the last few years [9]. tion for institutions that are initiating a LAAO program, given its The association between operatorś experience and LAAO procedural reproducible safety and efficacy results in centers unfamiliar with its results has been previously described. Indeed, several real-life studies prior device iteration and with lower procedural volume. However, have reported that increasing operatoŕs experience and hospital proce­ studies assessing longer-term outcomes following LAAO with the dural volume are linked with a lower rate of peri-procedural adverse Watchman FLX are warranted, in order to establish firm conclusions events, while the association with procedural success is more limited regarding the safety and efficacy of this device. [12–15]. Similarly, the latest registries with the Watchman 2.5 have reported growing technical success rates along with a declining inci­ 5. Limitations dence in peri-procedural complications, as compared to studies reporting on the initial experience with this device [3–4,9]. We acknowledge several limitations, inherent to the observational Thus, our findings suggest a lower learning curve for LAAO with the design of this study. Watchman FLX than that previously reported for the Watchman 2.5 Although information was entered prospectively by trained cardi­ device, with high procedural success rates achievable not only by more ologists involved in the procedure, we lacked a central core lab to assess expert operators, but also by those performing a lower volume of pro­ imaging and procedural data and event adjudication between the cedures and no prior experience with the Watchman 2.5 device. These different sites was not crosschecked. Of note, certain procedural details outcomes can be justified by several novel features of the Watchman FLX such as procedural time or transseptal puncture technique were un­ device that facilitate optimal device alignment and positioning, available for analysis. In addition, our study was circumscribed to the 6 I. Cruz-González et al. IJC Heart & Vasculature 38 (2022) 100941 Table 4 Table 5 Procedural and 7-day outcomes according to centerś experience with Watchman Procedural and 7-day outcomes according to the centerś procedural volume. 2.5 device. Centers with ≤ Centers with > Total p- No prior Prior Total p- 10 Watchman 10 Watchman sample value experience experience sample value FLX implants FLX implants (n ¼ Watchman 2.5 Watchman (n ¼ (n ¼ 92) (n ¼ 267) 359) (n ¼ 39) 2.5 359) Peri-procedural 1(1.1) 8(3.0) 9(2.5) 0.457 (n ¼ 320) complications(n,%) Peri-procedural 0 9(2.8) 9(2.5) 0.605 Death(n,%) 0 0 0 complications(n,%) Ischemic stroke/ 1(1.1) 0 1(0.3) 0.256 Death(n,%) 0 0 0 transient ischemic Ischemic stroke/ 0 1(0.3) 1(0.3) 0.891 attack /Systemic transient ischemic embolism(n,%) attack / Systemic Hemorrhagic stroke 0 0 0 embolism(n,%) (n,%) Hemorrhagic stroke(n, 0 0 0 Major bleeding 0 4(1.5) 4(1.1) 0.576 %) (BARC ≥ 3) (n,%) Major bleeding (BARC 0 4(1.3) 4(1.1) 0.630 Pericardial effusion 0 1(0.4) 1(0.3) 0.744 ≥ 3)(n,%) requiring Pericardial effusion 0 1(0.3) 1(0.3) 0.891 intervention(n,%) requiring Major vascular 0 1(0.4) 1(0.3) 0.744 intervention(n,%) complication(n,%) Major vascular 0 1(0.3) 1(0.3) 0.891 Significant procedural 0 1(0.4) 1(0.3) 0.744 complication(n,%) arrythmia(n,%) Ventricular arrythmia 0 1(0.3) 1(0.3) 0.891 Device-related 0.744 (n,%) complications(n,%) 0 0 0 Device-related 0.891 Device 0 1(0.4) 1(0.3) complications(n,%) 0 0 0 embolization Device embolization 0 1(0.3) 1(0.3) Device thrombosis Device thrombosis LAA leak post- 3(3.3) 7(2.6) 10(2.8) 0.488 LAA leak post- 1(2.6) 9(2.9) 10(2.8) 0.697 procedure(n,%) 2(2.2) 5(1.9) 7(2.0) procedure(n,%) 1(2.6) 6(1.9) 7(2.0) Mild(1–3 mm) 1(1.1) 2(0.8) 3(0.8) Mild(1–3 mm) 0 3(1) 3(0.8) Moderate(3–5 mm) 0 0 0 Moderate(3–5 mm) 0 0 0 Significant(>5mm) Significant(>5mm) Antithrombotic 0.021 Antithrombotic 0.004 therapy at therapy at discharge discharge(n,%) (n,%) None 6(6.5) 2(0.7) 9(2.5) None 1(2.6) 8(2.5) 9(2.5) Any antiplatelet 57(61.9) 181(67.8) 238 Any antiplatelet 34(87.2) 203(62.4) 237 therapy (66.3) therapy (66.1) Any anticoagulation 29(31.5) 84(31.5) 113 Any anticoagulation 4(10.3) 109(34.1) 113 (31.5) (31.5) Values:mean ± SD or n(%). Values:mean ± SD or n(%). BRAC: Bleeding Academic Research Consortium; DOAC:direct oral anticoagu­ BRAC: Bleeding Academic Research Consortium; ICE:intra-cardiac echocardi­ lants; LAA:left atrial appendage; LMWH:low-molecular weight heparin; VKA: ography; LAA:left atrial appendage. vitamin K antagonists. procedural phase and follow-up results including long-term device- Sources of funding related outcomes were not evaluated. Notwithstanding, prior studies with the Watchman FLX have reported low incidence of DRT and only No funding sources were employed for this study. small variations in the incidence of peri-device leaks, as compared to acute results assessed at the end of the procedure. Finally, the non­ Disclosures randomized design of the study and lack of a control group constitute a further limitation, and the possibility of patient selection bias cannot be Dr. I. Cruz-González is proctor for Lifetech Scientific [Shenzhen] Co., discarded. China and was funded by ISCIII (PI19/00658), co-funded by ERDF, “A way to make Europe” and Gerencia Regional Salud de CyL (GRS 3031/ 6. Conclusions A/19). Dr. A. Pérez de Prado is proctor for and has received Research grants from Boston Scientific [Marlborough, Massachusetts, US]. Dr R. The Watchman FLX attains high procedural success rates with Estévez-Loureiro and Dr. JM. Ruiz-Nodar are proctors for Boston Sci­ complete acute LAA sealing in a wide range of LAA anatomies, along entific [Marlborough, Massachusetts, US]. No other disclosures exist. with a very low rate of peri-procedural complications. Favorable peri- procedural outcomes following LAAO with the Watchman FLX are not Declaration of Competing Interest influenced by operatorś prior experience with the Watchman 2.5 device and are reproducible by operators at their initial experience with the The authors declare that they have no known competing financial Watchman FLX device. interests or personal relationships that could have appeared to influence the work reported in this paper. 7 I. Cruz-González et al. IJC Heart & Vasculature 38 (2022) 100941 [3] D.R. Holmes, S. Kar, M.J. Price, B. Whisenant, H. Sievert, S.K. Doshi, K. Huber, V. Y. Reddy, 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. 64 (1) (2014) 1–12. [4] L.V. Boersma, H. Ince, S. Kische, E. Pokushalov, T. Schmitz, B. Schmidt, T. Gori, F. Meincke, A.V. Protopopov, T. Betts, D. Foley, H. Sievert, P. Mazzone, T. De Potter, E. Vireca, K. Stein, M.W. Bergmann, Efficacy and safety of left atrial appendage closure with WATCHMAN in patients with or without contraindication to oral anticoagulation: 1- Year follow-up outcome data of the EWOLUTION trial, Heart Rhythm. 14 (9) (2017) 1302–1308. [5] X. Freixa, A. Tzikas, A. Basmadjian, P. Garceau, R. Ibrahim, The chicken-wing morphology: an anatomical challenge for left atrial appendage occlusion, J. Interv. Cardiol. 26 (2013) 509–514. [6] K. Korsholm, A. Samaras, A. Andersen, J.M. Jensen, J.E. Nielsen-Kudsk, The Watchman FLX Device: First European Experience and Feasibility of Intracardiac Echocardiography to Guide Implantation, JACC Clin. Electrophysiol. 6 (13) (2020) 1633–1642. [7] I. Cruz-González, K. Korsholm, B. Trejo-Velasco, et al., Procedural and short-term results with the new Watchman FLX left atrial appendage occlusion device, J. Am. Coll. Cardiol. Intv. 13 (2020) 2732–2741. [8] S. Kar, S.K. Doshi, A. Sadhu, R. Horton, J. Osorio, C. Ellis, J. Stone, M. Shah, S. R. Dukkipati, S. Adler, D.G. Nair, J. Kim, O. Wazni, M.J. Price, F.M. Asch, D. R. Holmes, R.D. Shipley, N.T. Gordon, D.J. Allocco, V.Y. Reddy, Primary Outcome Evaluation of a Next-Generation Left Atrial Appendage Closure Device: Results From the PINNACLE FLX Trial, Circulation. 143 (18) (2021) 1754–1762. [9] J.V. Freeman, P. Varosy, M.J. Price, D. Slotwiner, F.M. Kusumoto, C. Rammohan, C.J. Kavinsky, Z.G. Turi, J. Akar, C. Koutras, J.P. Curtis, F.A. Masoudi, The NCDR Left Atrial Appendage Occlusion registry, J. Am. Coll. Cardiol. 75 (13) (2020) 1503–1518. [10] A. Tzikas, D.R. Holmes, S. Gafoor, C.E. Ruiz, C. Blomström-Lundqvist, H.-C. Diener, Fig. 2. Peri-procedural complications. Peri-procedural major adverse events R. Cappato, S. Kar, R.J. Lee, R.A. Byrne, R. Ibrahim, D. Lakkireddy, O.I. Soliman, following LAAO with the Watchman FLX. P-value > 0.05 for all comparisons. M. Näbauer, S. Schneider, J. Brachman, J.L. Saver, K. Tiemann, H. Sievert, A. J. Camm, T. Lewalter, Percutaneous left atrial appendage occlusion: the Munich consensus document on definitions, endpoints and data collection requirements for Acknowledgments: clinical studies, EuroIntervention. 12 (1) (2016) 103–111. [11] V.Y. Reddy, D.N. Gibson, S. Kar, W. O’Neill, S.K. Doshi, R.P. Horton, M. Buchbinder, N.T. Gordon, D.R. Holmes, Post-Approval U.S. Experience With We would like to express our gratitude to Jose Santaursula Pastor for Left Atrial Appendage Closure for Stroke Prevention in Atrial Fibrillation, J. Am. the production of the figures in this work. Coll. Cardiol. 69 (3) (2017) 253–261. [12] L.V.A. Boersma, B. Schmidt, T.R. Betts, H. Sievert, C. Tamburino, E. Teiger, E. Pokushalov, S. Kische, T. Schmitz, K.M. Stein, M.W. Bergmann, Implant success Appendix A. Supplementary material and safety of left atrial appendage closure with the WATCHMAN device:peri- procedural outcomes from the EWOLUTION registry, Eur. Heart J. 37 (31) (2016) Supplementary data to this article can be found online at https://doi. 2465–2474. [13] J. Ledwoch, C. Krollmann, S. Staubach, M. Hug, H. Strohm, H. Mudra, Learning org/10.1016/j.ijcha.2021.100941. Curve Assessment for Percutaneous Left Atrial Appendage Closure With the WATCHMAN Occluder, J. Interv. Cardiol. 29 (2016) 393–399. References [14] I. Cruz-Gonzalez, A. Perez-Rivera, R. Lopez-Jimenez, J. Rodriguez-Collado, J. Martín-Moreiras, M. Cascon, A. Arribas, J.C. Gomez, A.O. Maree, C. Martin- Luengo, Significance of the learning curve in left atrial appendage occlusion with [1] M. Glikson, R. Wolff, G. Hindricks, J. Mandrola, A.J. Camm, G.Y.H. Lip, two different devices, Catheter. Cardiovasc. Interv. 83 (4) (2014) 642–646. L. Fauchier, T.R. Betts, T. Lewalter, J. Saw, A. Tzikas, L. Sternik, F. Nietlispach, [15] A.O. Badheka, A. Chothani, K. Mehta, N.J. Patel, A. Deshmukh, M. Hoosien, S. Berti, H. Sievert, S. Bertog, B. Meier, EHRA/EAPCI expert consensus statement N. Shah, V. Singh, P. Grover, G.T. Savani, S.S. Panaich, A. Rathod, N. Patel, on catheter-based left atrial appendage occlusion-an update, EuroIntervention. 15 S. Arora, V. Bhalara, J.O. Coffey, W. O’Neill, R. Makkar, C.L. Grines, T. Schreiber, (13) (2020) 1133–1180. L. Di Biase, A. Natale, J.F. Viles-Gonzalez, Utilization and adverse outcomes of [2] L. Asmarats, J. Rodés-Cabau, Percutaneous Left Atrial Appendage Closure: Current percutaneous left atrial appendage closure for stroke prevention in atrial Devices and Clinical Outcomes, Circ. Cardiovasc. Interv. 10 (11) (2017), https:// fibrillation in the United States: influence of hospital volume, Circ. Arrhythm. doi.org/10.1161/CIRCINTERVENTIONS.117.005359. Electrophysiol. 8 (1) (2015) 42–48. 8