Author + information
- Received January 31, 2014
- Revision received March 28, 2014
- Accepted April 10, 2014
- Published online September 1, 2014.
- Sergio Berti, MD∗∗ (, )
- Umberto Paradossi, MD∗,
- Francesco Meucci, MD†,
- Giuseppe Trianni, MD∗,
- Apostolos Tzikas, MD‡,
- Marco Rezzaghi, MD∗,
- Miroslava Stolkova, MD†,
- Cataldo Palmieri, MD∗,
- Fabio Mori, MD† and
- Gennaro Santoro, MD†
- ∗Adult Cardiology Department, Ospedale del Cuore G. Pasquinucci, Fondazione Toscana G. Monasterio, Massa, Italy
- †Interventional Diagnostic Department, Azienda Ospedaliera Universitaria Careggi, Florence, Italy
- ‡Interventional Cardiology Department, Interbalkan European Medical Center, Thessaloniki, Greece
- ↵∗Reprint requests and correspondence:
Dr. Sergio Berti, Adult Cardiology Department, Ospedale del Cuore, Fondazione Toscana G. Monasterio, Via Aurelia Sud, 54100 Massa, Italy.
Objectives This dual-center study sought to demonstrate the utility and safety of intracardiac echocardiography (ICE) in providing adequate imaging guidance as an alternative to transesophageal echocardiography (TEE) during Amplatzer Cardiac Plug device implantation.
Background Over 90% of intracardiac thrombi in atrial fibrillation originate from the left atrial appendage (LAA). Patients with contraindications to anticoagulation are potential candidates for LAA percutaneous occlusion. TEE is typically used to guide implantation.
Methods ICE-guided percutaneous LAA closure was performed in 121 patients to evaluate the following tasks typically achieved by TEE: assessment of the LAA dimension for device sizing; guidance of transseptal puncture; verification of the delivery sheath position; confirmation of location and stability of the device before and after release and continuous monitoring to detect procedural complications. In 51 consecutive patients, we compared the measurements obtained by ICE and fluoroscopy to choose the size of the device.
Results The device was successfully implanted in 117 patients, yielding a technical success rate of 96.7%. Procedural success was achieved in 113 cases (93.4%). Four major adverse events (3 cardiac tamponades and 1 in-hospital transient ischemic attack) occurred. There was significant correlation in the measurements for device sizing assessed by angiography and ICE (r = 0.94, p < 0.0001).
Conclusions ICE imaging was able to perform the tasks typically provided by TEE during implantation of the Amplatzer Cardiac Plug device for LAA occlusion. Therefore, we provide evidence that the use of ICE offered accurate measurements of LAA dimension in order to select the correct device sizes.
- atrial fibrillation
- intracardiac echocardiography
- intraprocedural imaging
- left atrial appendage occlusion
- transesophageal echocardiography
Stroke is the third cause of mortality in Western countries and the first cause of serious disability and morbidity (1). Patients with atrial fibrillation (AF) have an increased risk for stroke (2), ranging from 2% to >10% per year, depending on additional risk factors (3). As a result, AF is responsible for 15% to 20% of all ischemic strokes (4). The majority of ischemic strokes associated with AF are secondary to thromboembolism arising from the left atrial appendage (LAA). In a review of 23 studies in which the LAA was examined by autopsy, transesophageal echocardiography (TEE), or direct intraoperative inspection, intracardiac thrombus was identified in 17% of cases of nonvalvular AF, of which 91% were located in the LAA (5).
The efficacy of chronic anticoagulation therapy to prevent ischemic strokes in AF is well established (6), and current guidelines recommend an oral anticoagulation (OAC) regimen to prevent thromboembolism to all AF patients at risk of stroke (7,8). Management of warfarin, the most commonly administered OAC drug, is complicated by a narrow therapeutic window, the need for frequent monitoring, significant drug-to-drug interactions, and the increased risk of bleeding, which limit the use of warfarin for a substantial portion of these patients. Despite the introduction of novel OAC drugs, developed to overcome these disadvantages (9–11), a considerable number of AF patients do not receive OAC due to their bleeding risk or other contraindications. For these patients, the option of LAA occlusion by percutaneous implantation of an occlusive device has shown encouraging results as an alternative to OAC therapy (12,13).
The use of imaging techniques is of great relevance throughout all phases of the implantation procedure for LAA occlusion, specifically helping in the choice of the size of the device and in reducing the complication rate, which is still critical in the preliminary experience (13). The use of general anesthesia, typically required by TEE, can be avoided by application of intracardiac echocardiography (ICE). As previously reported by MacDonald et al. (14), this technique, used in LAA evaluation and closure for the first time by the investigators without TEE, reduced procedure time and can be performed safely under local anesthesia.
We present our dual-center experience regarding the utility and safety of ICE-guided percutaneous LAA occlusion as an alternative to implantation under TEE guidance.
Between January 1, 2009 and April 30, 2013, all consecutive patients undergoing LAA transcatheter occlusion procedure at the Ospedale del Cuore, Fondazione G. Monasterio in Massa, and at the Azienda Ospedaliera Universitaria of Careggi in Florence were enrolled in this study. Patients scheduled for percutaneous LAA occlusion had a history of nonrheumatic AF (chronic, paroxysmal, or persistent), a high stroke risk, and absolute contraindications to OAC. Contraindications to OAC that led to LAAC were extremely different in our population of 121 patients. In 42% of cases, patients had histories of major bleeding (47% of these were on OAC), mostly spontaneous or OAC-related intracranial bleedings. In 26.4% of cases, patients had histories of minor bleeding (56% of these were on OAC), mostly gastrointestinal bleedings. Also, in 9.1% of cases, there were histories of repeated thromboembolic events on warfarin. For the remaining cases, there were further contraindications to OAC, such as 2 cases of cerebral aneurysm, hematological disorders, carcinomas, high HAS-BLED (Hypertension, Abnormal renal and liver function, Stroke, Bleeding, Labile INRs, Elderly, Drugs or alcohol) scores, as well as others.
Characteristics of the occlusion device
All patients underwent TEE examination from 2 to 5 days before the procedure to exclude the presence of thrombus in the LAA and to assess the LAA anatomy, ostium and landing zone diameters, and length. In all cases, we were able to verify that the length of the LAA was ≥10 mm in order to correctly deploy the device with sufficient space. Patients were implanted with the Amplatzer Cardiac Plug (ACP) device or Amplatzer Cardiac Plug II (Amulet) (St. Jude Medical, Plymouth, Minnesota).
The ACP device is designed for immediate occlusion of the LAA and consists of a distal lobe and a proximal disk connected by an articulating waist (Figure 1). It is a self-expanding device made of a nitinol mesh and 2 patches of polyester sewn into both the lobe and the disk. The lobe of the device has 6 stabilizing wires that help to anchor the device in the LAA. The device is available in sizes between 16 and 30 mm (2-mm size increments, lobe size), and the proximal disk is 4 to 6 mm larger than the lobe. The lobe has a fixed length of 6.5 mm irrespective of device size. The ACP device is recommended to be oversized by about 2 to 3 mm with respect to the LAA dimensions. The device has a wide positional adaptability due to the connecting waist, which allows the disk of the system to self-orient from the lobe and completely occlude the ostium of the LAA.
The Amulet lobe has 12 stabilizing wires for the 16- and 18-mm devices, 16 stabilizing wires for the 20- to 25-mm devices, and 20 stabilizing wires for the 28- and 34-mm devices. Stabilizing wires help to anchor the device in the LAA. The device is available in sizes between 16 and 34 mm (2-mm size increments lobe size, for the 16- to 22-mm devices and 3-mm increments, lobe size, for the 22- to 34-mm devices). The proximal disk is 6 to 7 mm larger than the lobe. The lobe length is 6.5 mm for the 16- to 22-mm devices and 10 mm for the 25- to 34-mm devices. The Amulet device is recommended to be oversized by about 2 to 3 mm with respect to the LAA dimensions (Figure 1).
The implantation procedures were performed under local anesthesia and guided by fluoroscopy and ICE monitoring using the AcuNav catheter (Biosense Webster, Inc., Diamond Bar, California). This system is available in 2 sizes of 8- and 10.5-F shafts. It works with Sequoia, Cypress, or Aspen imaging systems, all of which are manufactured by Siemens Medical Solutions USA, Inc. (Malvern, Pennsylvania), and VIVID i (GE Healthcare, GE, Fairfield, Connecticut). The frequency of the transducer varies from 5.5 to 10 MHz. The catheter can be steered in 4 directions (anterior, posterior, left, and right). It also has color flow and spectral Doppler capabilities. Once the desired position of the catheter tip has been achieved, it can be set in that position with a lock mechanism, which is located on the catheter handle.
The ICE probe was positioned in the right atrium (RA) or in the coronary sinus (CS) position (Figures 2A and 2B). ICE guidance was used in the following ways: 1) to confirm the absence of LAA thrombus; 2) to identify the LAA landing zone dimension for the device sizing; 3) to guide the transseptal puncture; 4) to verify the delivery sheath position; 5) to confirm the location and stability of the device before and after release; and 6) to continuously monitor to detect any procedural complication, such as cardiac tamponade.
In the last 51 consecutive patients enrolled in Massa Ospedale del Cuore G. Pasquinucci, ICE was also used to assess the diameter of the LAA by measuring the transverse diameter 10 mm below the ostium, called the landing zone of the device. The same diameter was then measured by angiography by a different operator.
LAA occlusion: technique
Based on these measurements, the size of the device was chosen following the manufacturer’s recommended sheath size. Positioning and deployment of the device was performed under combined fluoroscopic and ICE guidance. The ideal lobe for sheath placement was decided on the basis of fluoroscopic images. The lobe of the device was placed inside the LAA, reaching the chosen landing zone while the disk was positioned in order to seal the LAA ostium and achieve complete exclusion of the appendage. Appropriate implantation was confirmed by a lobe positioned completely inside the LAA, aligned with the LAA landing zone and with a slight deformation of its body (tire-shape) (Figures 3A and 3B). The procedure has been previously described in detail by Berti et al. (15). Additional indicators for appropriate implantation included a clear separation between disk and lobe and a concave shape of the disk visible by fluoroscopy. Contrast fluid was injected through the sheath to confirm the position of the device and the effective-LAA occlusion. Device stability was verified by either ICE or fluoroscopy, as device embolization has been reported as a possible complication (12,13). In all patients, the proximal disk was kept on traction for at least 3 min to allow the device lobe to anchor within the LAA trabecular anatomy and to verify device stability (Figures 4A and 4B).
The analysis included data collected with respect to patient demographics and characteristics, procedural data, and complications that occurred during the implantation and subsequent hospitalization. Data were analyzed with emphasis on procedure-related aspects and acute treatment results. Specifically, the following endpoints were defined: 1) technical success: successful delivery and deployment of the device; and 2) procedural success: technical success and the absence of major adverse events during hospitalization. Major adverse events included death, stroke, embolism, pericardial or other major bleeding requiring intervention, device embolization, and major vascular complications.
All statistical analyses were conducted with the StatView statistical package (version 5.0.1, SAS Institute, Cary, North Carolina). Continuous variables are presented by mean ± SD (range) and event rates are expressed as percentages. Differences between LAA dimensions measured by ICE and fluoroscopy were evaluated by paired Student t test. Correlation between continuous variables was assessed by means of Pearson correlation coefficient. Statistical significance was defined as p < 0.05 (Figure 5).
A total of 121 patients with a history of nonrheumatic AF of at least 12 months, CHA2DS2VASc (Congestive heart failure [or left ventricular systolic dysfunction], Hypertension: blood pressure consistently >140/90 mm Hg [or treated hypertension on medication], Age ≥75 years, Diabetes mellitus, previous Stroke or TIA or thromboembolism, Vascular disease [e.g., peripheral artery disease, myocardial infarction, aortic plaque], Age 65 to 74 years, Sex category [e.g., female sex]) score ≥2 and absolute or relative contraindications to OAC were included. Mean age was 77 ± 7.6 years old (range 56 to 93 years) and 69 patients (57%) were men (Table 1).
The device was successfully implanted in 117 patients, yielding a technical success rate of 96.7%. LAA access was achieved by transseptal puncture in 118 of 121 (97.5%) of the cases, and in the remaining 3 cases (2.5%), a patent foramen ovale was used to enter the LAA.
In 51 patients, we compared the LAA ostium and landing zone diameters obtained by ICE and angiography. LAA dimensions measured by ICE and angiography differed minimally in all patients. The mean LAA landing zone for device diameter assessed by fluoroscopy was 21.0 ± 3.5 mm (range 14.0 to 29.4 mm), whereas by ICE it was 20.6 ± 3.3 mm (range 13.6 to 29.0 mm). The landing zones assessed by angiography and ICE were significantly correlated (Pearson correlation coefficient r = 0.94, 2-tailed test p < 0.0001).
Furthermore, ostium and landing zone measurements by ICE were significantly related to TEE (Pearson correlation coefficient r = 0.94, p < 0.0001; r = 0.72, p < 0.0001; for ostium and landing zone, respectively). More specifically, in 78.4% of the cases, the device selected based on ICE measurements was the same as the device selected based on TEE measurements. In 21.6% of cases, the device was 1 size larger, and in no cases was it smaller (Figure 6).
Of all 121 patients evaluated by TEE, the majority (67.7%) had a dual-lobe LAA anatomy that originated distally 10 mm from the ostium; a single-lobe anatomy was observed in 17.2% of the patients. The mean diameter of the device implanted was 24.0 ± 3.5 mm (range 16 to 34 mm). In 89.7% (105 of 117) of procedures, the first selected device was implanted; in 4 cases, the initial device was replaced by a smaller device; and in 8 other cases, a larger device was used. In all procedures, appropriate device positioning and effective LAA occlusion were accomplished in a single attempt. No obvious relationship was noticed between the number of attempts required and the LAA orifice diameter or the number of LAA lobes. The use of ICE together with angiography has, in all cases, provided us with an adequate guide, and in no cases did we need to refer back to the images acquired with TEE.
Among the 121 procedures, 4 adverse events and 3 minor complications occurred (Table 2). We had 4 cases of technical failure (3.3% of total). In 3 cases, the anatomy of the LAA was characterized by too many lobes and was not suitable for a device apposition. This was only discovered by angiography images, because they were not correctly detected by pre-procedural TEE. The cause of the last technical failure was an insufficiently sized landing zone. The most frequent adverse event was cardiac tamponade, which occurred in 3 patients. Two cardiac tamponades occurred 4 and 8 h, respectively, after intervention in patients whose procedures were prolonged due to many device-retrieval attempts; both complications were resolved by pericardiocentesis. The third cardiac tamponade occurred due to a LAA perforation and a consequent pulmonary artery laceration that was resolved surgically (16). In all these cases we had good ICE-derived images.
No cases of pericardial effusion without tamponade were observed in our cohort.
Another patient experienced a transient ischemic attack 24 h after the procedure. The brain computed tomography scan was negative for ischemic lesion, and the event spontaneously resolved within a few hours. The patient was discharged without any sequela. Minor complications included 2 cases of femoral hematomas. In 1 case, the femoral hematoma was on left side, the ICE access side. This complication did not require a blood transfusion or reintervention, only a compression of a major duration without a prolonged hospitalization.
Accounting for the 4 major adverse events, the procedural success rate was 93.4% (113 of 121 cases).
The use of procedural echocardiography is an essential requirement for safe and successful percutaneous LAA occlusion. Without disputing the value of a pre-procedural TEE examination to explore the relevant cardiac anatomy and exclude thrombi, our results suggest that ICE is an acceptable alternative to TEE during implantation of the ACP device.
ICE is an imaging technique applied in various interventional cardiac procedures, including catheter ablation of AF, percutaneous closure of atrial and ventricular septal defects, and mitral valvuloplasty (17,18). For a series of 10 patients, ICE imaging during percutaneous LAA occlusion was previously reported to be optimal with the ICE probe positioned in the coronary sinus (19). In 28 of 51 cases, we integrated the images with views both from the CS and the RA. While the ICE views from the CS provided good images, in our experience, this is not fundamental because we achieved satisfactory images for all procedural aspects from the standard RA position. Although ICE may result in an inferior image quality compared with that of TEE, in our experience this is not affecting the safety and technical success of the implantation procedure, provided that key information is recorded during the pre-procedural TEE examination. In the case of both pre-procedural TEE and periprocedural ICE, it is fundamental to find the measures of the LAA ostium and landing zone. It is not easy to evaluate the full length of the appendage because of the complex geometry of the LAA; however, a critical point is to verify that the minimum length from the neck must be ≥10 mm in order to have sufficient room for the correct deployment of the device.
ICE-guided percutaneous LAA occlusion is feasible with similar results to those reported for TEE-guided implantation. Technical success rates for percutaneous LAA occlusion are reported to be 90% and higher (13,20) compared with 96.7% in this study. Typical rates for major complications, including significant pericardial effusion, tamponade and periprocedural stroke, are reported to be between 3.7% and 7.7% (13,20). In this study, adverse events occurred in 3.4% of the procedures, including 3 pericardial tamponades and 1 transient ischemic attack. Another critical point is the presence of peridevice leaks after device implantation. Angiography and TEE are undoubtedly more precise than ICE in monitoring the presence of any peridevice leaks. In our experience, it has been established that small leaks that were visible only by angiography did not present a need for device substitution. However, leaks that were visible both with angiography and ICE were motivation for device substitution.
As stated by some investigators, there is some experience of fluoroscopy-only-guided LAAC that has demonstrated the reliability of a procedure using local anesthesia and fluoroscopy alone (21). In our opinion, purely fluoroscopic guidance in LAAC has significant limitations. Hence, complications may result from suboptimal catheter handling and device performance from undetected thrombus formation, bleeding, and wall disruption. Especially for learning curve centers, we recommend the use of fluoroscopy integrated with echocardiographic imaging.
Concerning the periprocedural assessment of the appropriate device size, a high level of agreement was found between results from fluoroscopy and ICE. The initially selected device was only exchanged for a differently sized device in 10.3% of the cases. This compares well with results reported for the same device by operators using TEE (13,22).
Application of TEE and endotracheal intubation may contribute to complications such as injuries and bleeding of gastroesophageal and respiratory tract, laryngospasm and bronchospasm requiring additional intervention, and/or prolonged hospitalization. The use of ICE eliminates the need for general anesthesia, which is beneficial for all, especially older patients. Moreover, as shown in our study, left femoral vein hematoma is the only ICE-related complication, a minor complication not requiring any blood transfusion reintervention or prolonged hospitalization. With regard to the percutaneous closure of interatrial communications, ICE, compared with TEE, has been reported to result in less procedural discomfort for the patient and reduced procedural time and ionizing radiation exposure (23) and to avoid the obstruction of fluoroscopic viewing. Although not assessed in our study, it seems reasonable to assume that similar benefits exist for percutaneous LAA occlusion.
The use of ICE instead of TEE does not require the presence of an anesthesiologist, and therefore the amount of global exposure to ionizing radiation is also reduced for the operators involved. In periprocedural use of TEE or ICE, 2 operators are always necessary (1 for the implantation and 1 for the imaging). The safety issue is the proximity of the TEE operator to the x-ray tube and the lack of x-ray screen protection in that position. In periprocedural ICE use, both operators are on the same side as the patient behind the x-ray screen protection.
It is questionable whether the additional ICE probe increases the overall cost of the procedure. That cost is probably balanced by the added cost of the TEE-experienced echocardiographer, the anesthesiology team, and the longer catheterization lab occupancy. Furthermore, the use of local anesthesia may reduce direct medication and personnel costs (24,25), as well secondary costs related to TEE complications.
In our patients, the visualization of the LAA was feasible with the ICE using the RA and CS positions, and the measurements obtained by angiography and ICE were significantly correlated.
The LAA dimension is essential for selecting the correct size of the device, and we observed that the accuracy of ICE is comparable with angiography in evaluating LAA measurements. ICE may represent a valid support for imaging guidance during the percutaneous LAA closure procedure.
Our study was a dual-center observational study without a control group, and our results may not be extrapolated to other LAA occlusion devices and other types of ICE catheters. Further investigations are required to fully characterize the application of ICE during percutaneous LAA occlusion, including its impact on safety, the length of learning curve, procedural success, and healthcare costs.
Although accepted as the guiding tool for percutaneous LAAC device closure procedures and for electrophysiological catheter-based ablations, randomized multicenter trials on ICE in this context are lacking. In our experience, ICE represents a safe and useful ultrasound option for guiding the LAA transcatheter occlusion procedure and for preventing short- and mid-term complications having the advantage over TEE of not requiring the support of general anesthesia and anesthesiology. These factors provide direct benefits to patients. Compared with fluoroscopy, ICE also showed a good accuracy in assessing LAA dimension in order to select the correct device sizes.
The authors thank Karin Joan Tyack for her linguistic support.
Dr. Tzikas is a consultant for St. Jude Medical. Dr. Berti is a proctor for both St. Jude and Edwards Lifesciences. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- Amplatzer Cardiac Plug
- atrial fibrillation
- coronary sinus
- intracardiac echocardiography
- left atrial appendage
- LAA Closure
- oral anticoagulant(s)
- right atrium
- transesophageal echocardiography
- Received January 31, 2014.
- Revision received March 28, 2014.
- Accepted April 10, 2014.
- American College of Cardiology Foundation
- Sacco R.L.,
- Benjamin E.J.,
- Broderick J.P.,
- et al.
- Wolf P.A.,
- Abbott R.D.,
- Kannel W.B.
- Camm A.J.,
- Kirchhof P.,
- Lip G.Y.,
- et al.,
- for the European Heart Rhythm Association, European Association for Cardio-Thoracic Surgery
- Camm A.J.,
- Lip G.Y.,
- De Caterina R.,
- et al.,
- for the ESC Committee for Practice Guidelines
- Hijazi Z.M.,
- Shivkumar K.,
- Sahn D.J.
- Reddy V.V.,
- Holmes D.,
- Doshi S.K.,
- Neuzil P.,
- Kar S.
- Meerkin D.,
- Butnaru A.,
- Dratva D.,
- Tzivoni D.