Author + information
- Received April 26, 2019
- Revision received August 5, 2019
- Accepted August 14, 2019
- Published online February 3, 2020.
- Kasper Korsholm, MDa,
- Sergio Berti, MDb,
- Xavier Iriart, MDc,
- Jacqueline Saw, MDd,
- Dee Dee Wang, MDe,
- Hubert Cochet, MD, PhDc,
- Danny Chow, MDf,
- Alberto Clemente, MDg,
- Ole De Backer, MD, PhDh,
- Jesper Møller Jensen, MD, PhDa and
- Jens Erik Nielsen-Kudsk, MD, DMSca,∗ ()
- aDepartment of Cardiology, Aarhus University Hospital, Aarhus, Denmark
- bDepartment of Cardiology, Fondazione CNR Regione Toscana, Massa, Italy
- cBordeaux University Hospital (CHU), IHU Liryc, Electrophysiology and Heart Modeling Institute, Fondation Bordeaux Université, Bordeaux, France
- dDivision of Cardiology, Vancouver General Hospital, Vancouver, British Columbia, Canada
- eCenter for Structural Heart Disease, Division of Cardiology, Henry Ford Health System, Detroit, Michigan
- fDepartment of Medicine and Geriatrics, Princess Margaret Hospital, Hong Kong, China
- gDepartment of Radiology, Fondazione CNR Regione Toscana, Massa, Italy
- hDepartment of Cardiology, Heart Center, Rigshospitalet, Copenhagen, Denmark
- ↵∗Address for correspondence:
Dr. Jens Erik Nielsen-Kudsk, Department of Cardiology, Aarhus University Hospital, Palle Juul-Jensens Boulevard 99, DK-8200 Aarhus N, Denmark.
• Cardiac CT is increasingly used to plan transcatheter LAAO.
• High-quality multiplanar and 3D imaging allow accurate device sizing and efficient LAAO.
• A standardized protocol for cardiac CT is paramount for a high-quality result.
• Pre-procedural cardiac CT may improve the accuracy, efficacy, and safety of LAAO.
Transcatheter left atrial appendage occlusion is an increasingly used alternative to oral anticoagulation in selected patients with atrial fibrillation. Pre-procedural imaging is a prerequisite to a successful intervention, with transesophageal echocardiography as the current gold standard. However, cardiac computed tomography offers improved imaging with high-quality multiplanar and 3-dimensional reconstructed images. Nevertheless, the lack of a standardized imaging protocol has slowed the adoption of cardiac computed tomography into clinical practice. On the basis of current research and expert consensus, this paper provides a protocol for the preparation, acquisition, and interpretation of cardiac computed tomographic imaging in pre-procedural planning of left atrial appendage occlusion.
- CT scanner requirements
- left atrial appendage thrombus detection and measurements
- contrast injection
- patient preparation
Atrial fibrillation (AF) is an increasing health care concern, with a projected incidence of up to 215,000 per year in Europe by 2030 (1). The risk for thromboembolism associated with AF is about 5% per year, with substantially higher mortality and morbidity in case of AF-related stroke compared with non-AF-related stroke (2). Left atrial appendage occlusion (LAAO) is a nonpharmacological therapy for stroke prevention in selected patients with AF and increased risk for thromboembolism (3). Two randomized trials and multiple observational registries have demonstrated the efficacy and safety of LAAO (4–10), and multiple trials are ongoing.
The anatomy of the left atrial appendage (LAA) is highly variable; hence, pre-procedural imaging is paramount to guide the operator in accurate sizing, device selection, and planning of the intervention. Two-dimensional transesophageal echocardiography (TEE) is currently considered the gold-standard imaging modality for LAAO. Observational studies report larger LAA dimensions using 3-dimensional (3D) imaging (11–15). A small randomized trial highlighted that 3D cardiac computed tomography (CT) improved pre-procedural device selection accuracy compared with 2-dimensional TEE, resulting in shorter procedure times (16). Despite apparent advantages of CT, universal adoption of pre-procedural cardiac CT has been slowed by the lack of a standardized imaging protocol, unfamiliarity with image and software manipulations, and concerns regarding radiation exposure, contrast use, and cost.
This consensus document has been developed by experienced European and American LAAO operators and cardiac computed tomographic imagers who regularly use cardiac CT as part of their work flow for percutaneous LAAO. On the basis of a review of currently available published research and expert consensus, this document provides recommendations for the preparation, acquisition, reconstruction, and interpretation of cardiac CT before LAAO procedures. The document provides a standardized, systematic protocol for physicians aiming to incorporate 3D cardiac CT into their LAAO work flow.
Cardiac CT in Planning LAAO
The LAA has a complex and highly heterogenous anatomy. It is extensively trabeculated, often multilobed and thin-walled. The anatomic complexity promotes slow flow, blood stasis, and thrombus formation, in particular in the presence of AF with lack of atrial systole. Additionally, the complex anatomy makes transcatheter LAAO challenging, underlining the importance of accurate pre-procedural imaging to optimize the accuracy, efficacy, and safety of the procedure. Pre-procedural cardiac CT should be performed with the objective of determining the anatomic feasibility of the LAAO procedure, to provide accurate sizing of the LAA dimensions for device selection, peri-procedural planning, and exclusion of LAA thrombus (Central Illustration).
Additionally, 3D volume rendering provides a roadmap for transseptal puncture and optimal C-arm angulations for device implantation (12), which may reduce the amount of contrast used during the procedure (17).
Available Computed Tomographic Scanner Systems and Minimum Requirements
The 4 major vendors are Siemens Medical Systems, GE Healthcare, Philips Medical Systems, and Canon Medical (former Toshiba Medical). Three vendors have a similar approach in their premium scanner cardiac technique, with a volume scan covering the cardiac anatomy in 1 rotation or 1 heartbeat, which is being used by Canon Medical and GE Healthcare, while Philips does not permit full heart coverage (coverage 8 cm), and a second scan is necessary for full coverage (Jog Scan). Siemens uses a dual-source design to offer high-speed spiral scanning through fast tube rotation and high patient table movement speed (Flash scan high-pitch acquisition). Helical or spiral scanning with retrospective electrocardiographic (ECG) gating or sequential scanning with prospective ECG gating is available in all scanners. In addition, helical or spiral scanning with prospective ECG gating is possible but not provided by all scanner systems. A 64-slice computed tomographic scanner equipped with a cardiac package is regarded as the minimum requirement to obtain sufficient images for LAAO. Newer scanners provide more sensitive detectors, faster gantry speed rotation, and improved iterative and model-based reconstruction software. These technological improvements enable high-quality diagnostic cardiac CT images, with very low radiation exposure. Hence, a minimum 128-slice computed tomographic scanner is recommended to increase image quality and minimize radiation exposure to patients. All vendors offer these minimum requirements in their current portfolios.
Patient Preparation for CT
An inherent advantage of cardiac CT is the ease and efficiency of this noninvasive imaging acquisition; hence, requirements for fasting and potential risks for esophageal injury from TEE are obsolete, and sedation is redundant. However, a proper patient education and image acquisition protocol is critical to a reproducible and dependable dataset acquisition.
Patient preparation before CT
Patients should arrive for their cardiac computed tomographic scans in a nonfasting state. Research has shown LAA size to vary on 2-dimensional TEE on the basis of different loading conditions of the left atrium (18). We recommend that patients maintain their regular daily oral fluid intake prior to arrival at their CT appointments to mimic their real-life LAA loading conditions. Upon arrival to the CT appointment, 250-ml water oral intake may be considered to ensure that patients are not dehydrated. For patients on dialysis, ideal time of CT scanning would be prior to the next dialysis session in order to achieve adequate loading conditions. In rare circumstances, patients may be fasting for other examinations on the day of cardiac CT. In those situations, hypovolemia should be corrected prior to the scan, and intravenous 5 to 10 ml/kg fluid infusion may be considered.
Caffeinated drinks such as tea, coffee, and cola as well as smoking prior to the scan should be avoided, in order not to affect the patient’s heart rate.
Patient preparation upon arrival to computed tomographic scanner
Patients should be positioned in a supine position, at the isocenter of the scanner. The arms should be raised above the head and the legs positioned comfortably using a leg support. The ECG electrodes are placed outside the scan range, and a 16- to 18-gauge intravenous needle is recommended in the right antecubital vein.
A thorough walk-through of the different steps of the scan with the patient is mandatory to avoid motion artifacts. Information must include potential side effects of contrast infusion (heat sensation and urinary urge). Breath-hold training is compulsory, with multiple shallow breath-holds performed prior to the scan to ensure patient cooperation, and monitoring of heart rate variability for final scan protocol optimization.
Patient considerations: Renal function, body mass index, and medication
Kidney function should be known prior to contrast administration in accordance with local guidelines. In general, the risk for contrast-induced nephropathy is considered very low in patients with glomerular filtration rates >45 ml/min. Hence, this threshold is regarded as safe for contrast administration (19). A glomerular filtration rate between 30 and 45 ml/min is not considered a contraindication to cardiac CT; however, dehydration should be avoided, and contrast volume should be minimized to reduce the risk for contrast-induced nephropathy. In patients with glomerular filtration rates <30 ml/min, the indication for cardiac CT should be carefully evaluated, taking several other factors into consideration, such as age, concomitant medication, previous reaction (renal dysfunction) to contrast administration, and declining or stable renal function. A nephrologist consultation may be recommended, and alternative LAA and cardiac imaging with TEE should be considered. Routine follow-up of renal function is central in patients with renal impairment and at risk for contrast-induced nephropathy.
A high body mass index (BMI) may impair image quality; however, no upper threshold can be defined. In accordance with the ALARA (as low as reasonably achievable) principles, an individual assessment must be undertaken case by case.
Patients without renal disease should continue their daily medications. Nephrotoxic medications and metformin therapy are recommended to be held in patients with renal impairment, in accordance with local guidelines.
Technical Protocol for Cardiac CT
Prospective versus retrospective ECG gating
Both prospective and retrospective ECG gating have been used for pre-procedural imaging acquisition (20). Retrospective ECG gating is the continuous acquisition of raw radiographic images synchronized with the electrocardiogram (ECG pulsing). The raw data are reconstructed post hoc during a specified cardiac phase. Prospective ECG gating is the online interpretation of the electrocardiogram, and on the basis of the R-R interval in the previous cycle, acquisition of data is triggered at the desired cardiac phase, defined a priori.
Prospective ECG gating is more sensitive to cardiac arrhythmias and does not allow post hoc reconstruction of cardiac phases, but the radiation exposure is significantly reduced compared with retrospective ECG gating. Consequently, it is recommended to use a prospective electrocardiographically gated acquisition to respect the ALARA principles. Images should be obtained in a phase corresponding to 30% to 60% of the R-R interval (21,22); preferably at a narrow pre-defined R-R interval within this range to reduce radiation exposure.
The acquisition technique depends on the scanner system used (Table 1). A topogram is mandatory to adjust the scanning volume in the craniocaudal direction to secure coverage of the heart from the carina to the diaphragm (10 to 16 cm). To determine the arrival of contrast and trigger the proper time of the scan sequence, most centers use a bolus-tracking technique (i.e., sureStart/CARE Bolus/SmartPrep), with a region of interest placed in the left atrium or ascending aorta (13,20,23–25). Some use a test bolus technique consisting of monitoring contrast transit times using a test bolus. This method is equally recommended, provided it does not expose patients with borderline renal function to increased contrast volumes. The tube potential should be set between 80 and 140 kV depending on BMI. The cutoff BMI value varies depending on detector design and sensitivity of the computed tomographic system. Generally, patients with BMIs >27 to 30 kg/m2 may require tube potential >100 kV (23,24), whereas others use an optional correction to the automated software based on the actual weight of the patient (20). The tube current should be adjusted to BMI following the settings recommended by each vendor and may be modulated over the cardiac cycle to save dose by using automated current modulation, a technology available on most scanner systems. The scan should be electrocardiographically gated, with a prospective protocol aiming for a cardiac phase within 30% to 60% of the R-R interval using volume, spiral, or step-and-shoot acquisition according to the scanner used. Slice thickness should be between 0.5 and 1.0 mm. Specific protocol modifications may apply to different scanner systems, and a proposed protocol setup is displayed in Table 2.
Contrast injection protocol
A proper contrast protocol is paramount to obtain a good imaging result. A contrast medium concentration of 350 mg iodine/ml is preferred, with volumes between 40 and 100 ml depending on patient weight and renal function. A dual-head power injector is recommended, with a saline chaser following the contrast medium. Various injector schemes are available; a biphasic protocol using 5 to 7 ml/s iodine contrast followed by a saline chaser of 50 ml at the same injection rate usually reduces streak artifacts by leaving the right heart cavities washed out. A triphasic protocol using 5 to 7 ml/s iodine contrast followed by a contrast/saline mixture and saline chaser to push the contrast column toward the right atrium keeps opacification in the right heart cavities, thus enabling assessment of the septum. However, we recommend using a biphasic contrast injection protocol.
Pharmacological heart rate control and nitrate administration
No clear upper limit for heart rate can be determined, and the limit is likely dependent on the scanner system used. In general, heart rate control is not a requirement with the current scanner technologies, which provide better temporal resolution for LAA imaging. However, pharmacological heart rate control may be considered in patients with very high heart rates, especially at centers using older computed tomography scanner systems with reduced temporal resolution.
In general, sublingual nitrates are not required for cardiac CT for LAAO, unlike for coronary artery imaging, for which they are routinely used to increase the visibility of the circumflex artery (20).
Delayed imaging acquisition: Thrombus rule-out
At centers performing pre-procedural cardiac CT with the aim of excluding LAA thrombus, it is highly recommended to perform delayed imaging for reliable thrombus exclusion and to avoid confirmatory TEE. Delayed imaging is the application of a 2-phase scan protocol, in which a second set of images are acquired following a short time delay (30 to 180 s) from the initial scan. This protocol adaptation aims to increase the diagnostic ability to detect LAA thrombus. A meta-analysis of contemporary studies using cardiac CT for detection of LAA thrombus, and TEE as the reference, showed a weighted mean sensitivity of 100% and a negative predictive value of 100%. Specificity ranged from 67% to 100% across studies and the positive predictive value from 12% to 100%. However, with delayed imaging scan, specificity consistently increased to between 98% and 100% (26,27). We recommend using a 60-s delay from contrast peak detected by bolus tracking. The delayed scan can be performed using a lower tube voltage to enhance x-ray interaction with the iodine media and to reduce the radiation delivered to the patient.
Alternative protocol adaptations to exclude LAA thrombus have been investigated. For example, a double-contrast single-phase acquisition has shown promising results but doubles the contrast load (28). Likewise, reports of scans performed with the patient in a prone position suggest increased diagnostic accuracy in patients with persistent filling defects (27). However, these protocol modifications are not generally recommended on the basis of the limited evidence and work flow challenges required in repositioning patients.
The ALARA principles should be followed. Scan parameters should be tailored on the basis of the patient’s weight, BMI, and heart rhythm. The latest computed tomographic scanners include a suite of automated control systems to aid the operator in keeping the radiation dose low and maximizing the probability of high-quality, motion-free images. However, operators should be familiar with radiation-reducing techniques to perform a manual adjustment (Table 3). Dose modulation systems (XYZ modulation, ECG modulation), iterative reconstruction kernels, tube current modulation, ECG gating techniques, scan length, and patient positioning on the table may reduce radiation exposure.
Pre-LAAO Cardiac Computed Tomographic Analysis
Multiple software packages are available to assist in imaging post-processing, such as 3mensio software (Pie Medical Imaging, Bilthoven, the Netherlands), Brilliance (Philips Healthcare, Eindhoven, the Netherlands), OsiriX software (Pixmeo, Bernex, Switzerland), syngo.via (Siemens Healthcare, Erlangen, Germany), Mimics (Materialise, Leuven, Belgium), and Vitrea (Toshiba Medical Systems, Zoetermeer, the Netherlands). Although each has different structural heart packages, all generate 3D multiplanar reconstructions.
It is recommended that the LAAO operator become familiar with the analysis and interpretation of pre-procedural cardiac CT to actively participate in the planning of the intervention. An overview of the items to assess and report during planning is presented in Table 4.
The initial exclusion of LAA thrombus can be performed on the basis of the axial views of the appendage (Figure 1). Contrast filling defects may represent slow contrast mixing due to low flow velocities, hence a comparison between the early and the delayed images will determine whether the filling defect represents a thrombus. A filling defect visible in the early acquisition but not the late acquisition indicates slow flow. Contrast filling defects present in both the early and delayed image acquisition most likely represent thrombus and should trigger additional imaging to verify or exclude LAA thrombus (i.e. TEE, intracardiac echocardiography).
Anatomic suitability for device implantation can be determined on the basis of a morphological assessment of the appendage in the volume-rendered 3D reconstruction. Here, the spatial relationship can be appreciated (Figure 2). Furthermore, feasibility of LAAO can be determined on the basis of the double-oblique multiplanar views of the LAA (Figure 3, Online Videos 1 and 2). From the axial 4-chamber view of the heart, a 2-chamber long-axis view is obtained by tilting the coronal plane (Figures 3D to 3F), corresponding to a right anterior oblique angulation of about 30°. In the 2-chamber view, the crosshairs are positioned at the base of the LAA, rotating it to align with the left upper pulmonary vein ridge and the circumflex coronary artery and verifying the perpendicular plane in the short-axis view (Figures 3G to 3I). A third en face projection of the LAA ostium is presented (Figure 3G). To evaluate the landing zone of the LAA, the same procedure is performed from the 2-chamber long-axis view. The position and definition of the landing zone are different for each of the available LAAO devices, but the methodological approach to define the landing zone is similar. The crosshairs are positioned at the intended position, and the axes are aligned with the LAA wall in both the long-axis and short-axis views to generate an en face view of the landing zone. Here, the maximum and minimum diameter can be obtained, along with perimeter- and area-derived diameters (Figure 4). Each device manufacturer has its own guidelines for determining which size of device to use, and currently the device sizing is based on the maximum diameter of the landing zone, with 3- to 6-mm device oversizing. However, studies are warranted to investigate the optimal use of maximum perimeter- or area-derived diameters in different anatomies and with different devices. The depth of the appendage should be obtained, along with the length of the dominant lobe (Figure 4⇓, Online Videos 1 and 2). The minimal sufficient length and depth required for device implantation are indicated in the manufacturer guidelines. Additionally, the optimal C-arm angulation for intraprocedural guidance can readily be obtained from the post-processing software. A thorough assessment of the atrial septal anatomy may facilitate a safe and optimal transseptal crossing.
Finally, an assessment of the cardiac valves, coronary artery anatomy, and coscanned lung segments should be considered depending on the clinical situation, patient’s comorbidities, and image quality.
Future perspectives on cardiac CT
Pre-procedural planning using cardiac CT is under continuous development, and its potential utility is constantly expanding. The introduction of 3D printing in ambiguous and challenging anatomies has shown promising results in case series (12) (Figure 5). The use of computational modeling may enhance the ability to predict optimal device size and position of the device to obtain complete sealing (Figure 6). Fusion or overlay imaging with cardiac computed tomographic images transferred onto real-time intraprocedural fluoroscopic angiograms has huge potential in guiding the interventionalist (Figure 7⇓⇓). In turn, some post-processing workstations have implemented simulation of the projection and angulation between the transseptal puncture site and LAA, to help determine the optimal transseptal puncture site and choice of delivery sheath shape.
The image acquisition of the LAA by cardiac CT requires training and knowledge, and preferably all institutions entering this field should have experience with performing cardiac CT for other cardiac diseases or interventions. To gain the optimal advantages of pre-procedural planning using cardiac CT, establishment of the work flow should be performed in close collaboration between imagers and operators.
This document provides an overview of the requirements needed to perform cardiac computed tomographic acquisition along with a practical set of recommendations for the preparation of patients, technical acquisition of imaging data, and post-processing for optimal use for LAAO.
Cardiac CT is a noninvasive, highly efficient, and valuable imaging modality for pre-procedural planning of LAAO. The isotropic imaging acquisition provides high-quality 3D images with multiplanar reconstruction, allowing optimal planning of a complex procedure such as LAAO. Cardiac CT has the potential to become the new gold standard for pre-procedural planning of LAAO.
The authors thank Dr. Ciobotaru Vlad (3DHeartModeling, Department of Cardiac Imaging, Hospital Privé Les Franciscaines, Nimes, France), who kindly provided pictures of the 3D printed models.
Dr. Korsholm has received a speaking honorarium from Abbott. Dr. Berti is a proctor for Abbott and Edwards Lifesciences. Dr. Iriart is a proctor for Boston Scientific and Abbott. Dr. Saw has received unrestricted research grant support from AstraZeneca, Abbott, Boston Scientific, and Servier; has received speaking honoraria from AstraZeneca, Abbott, Boston Scientific, and Sunovion; and is a consultant and/or proctor for Boston Scientific, AstraZeneca, and Abbott. Dr. Wang has received research grant support from Boston Scientific; and is a consultant for Edwards Lifesciences, Boston Scientific, and Materialise. Dr. De Backer has been a consultant for Boston Scientific and Abbott. Dr. Møller Jensen has received a speaking honorarium from Bracco Imaging. Dr. Nielsen-Kudsk is a consultant for Boston Scientific; and is a consultant and proctor for Abbott. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- atrial fibrillation
- as low as reasonably achievable
- body mass index
- computed tomography
- left atrial appendage
- left atrial appendage occlusion
- transesophageal echocardiography
- Received April 26, 2019.
- Revision received August 5, 2019.
- Accepted August 14, 2019.
- 2020 American College of Cardiology Foundation
- Marini C.,
- De Santis F.,
- Sacco S.,
- et al.
- Kirchhof P.,
- Benussi S.,
- Kotecha D.,
- et al.
- Reddy V.Y.,
- Doshi S.K.,
- Kar S.,
- et al.
- Landmesser U.,
- Tondo C.,
- Camm J.,
- et al.
- Landmesser U.,
- Schmidt B.,
- Nielsen-Kudsk J.E.,
- et al.
- Boersma L.V.,
- Ince H.,
- Kische S.,
- et al.
- Holmes D.R.,
- Kar S.,
- Price M.J.,
- et al.
- Wang D.D.,
- Eng M.,
- Kupsky D.,
- et al.
- Rajwani A.,
- Nelson A.J.,
- Shirazi M.G.,
- et al.
- Chow D.H.,
- Bieliauskas G.,
- Sawaya F.J.,
- et al.
- Saw J.,
- Fahmy P.,
- Spencer R.,
- et al.
- Eng M.H.,
- Wang D.D.,
- Greenbaum A.B.,
- et al.
- Spencer R.J.,
- DeJong P.,
- Fahmy P.,
- et al.
- Korsholm K.,
- Jensen J.M.,
- Nielsen-Kudsk J.E.
- Staab W.,
- Goth S.,
- Sohns C.,
- et al.
- Cochet H.,
- Iriart X.,
- Sridi S.,
- et al.
- Saw J.,
- Lopes J.P.,
- Reisman M.,
- McLaughlin P.,
- Nicolau S.,
- Bezerra H.G.
- Jaguszewski M.,
- Manes C.,
- Puippe G.,
- et al.
- Romero J.,
- Husain S.A.,
- Kelesidis I.,
- Sanz J.,
- Medina H.M.,
- Garcia M.J.
- Hur J.,
- Kim Y.J.,
- Lee H.-J.,
- et al.
- Central Illustration
- Cardiac CT in Planning LAAO
- Available Computed Tomographic Scanner Systems and Minimum Requirements
- Patient Preparation for CT
- Technical Protocol for Cardiac CT
- Radiation Exposure
- Pre-LAAO Cardiac Computed Tomographic Analysis