Left atrial appendage occlusion with the Amulet device: to tug or not to tug?

Abstract

Purpose

Left atrial appendage occlusion (LAAO) involves a “tug test,” in which implanters pull on the device delivery cable to ensure stable occluder placement. The aim of this study was to evaluate the recommendation to perform the tug test, by comparing forces exerted on the device during deployment and subsequent tug test. A secondary objective was to simulate forces experienced on left atrial appendage tissue by placement of a 20-mm device.

Methods

The AMPLATZER™ Amulet™ device was used for occlusion. A force transducer recorded forces in the delivery cable during deployment and tug test in 23 patients. Four patients were excluded due to improper transducer placement or technical errors in data collection. For a 20-mm device, the force imparted on the circumferential contact with left atrial appendage wall tissue was simulated in a computational model, using the measured externally applied forces as inputs.

Results

For devices < 25-mm in diameter, disc deployment force (mean ± standard deviation) was 1.72 ± 0.43 N, and tug force was 1.01 ± 0.59 N. For devices ≥ 25 mm in diameter, disc deployment force was 2.96 ± 0.57 N, and tug force was 1.04 ± 0.24 N. The increase in disc deployment force compared with tug test force was statistically significant for small devices (< 25 mm; p = 0.049) and large devices (≥ 25 mm; p < 0.001).

Conclusions

Increased force applied on the AMPLATZER™ Amulet™ device during disc deployment compared with during tug test was statistically significant, suggesting that the tug test is redundant in most cases for checking device stability.

Appendix

To solve the computational LAA model, the mechanical behavior of the device was assessed as experimental stress-strain data. Assuming linear elastic properties, a computational model of the LAA with implanted disc was used to evaluate the relationship between the measured forces in the delivery cable and those exerted at the LAA tissue. Figure 5 shows the workflow used.

Fig. 5

Summary flowchart of the steps taken to relate externally measured forces with internally experienced forces on LAA tissue

Stress-strain data were assessed through uniaxial tensile testing (Bose® ElectroForce® 3200 Series II, Mountain View, USA); 3D LAA geometry and mesh were created using Simpleware ScanIP, as noted in the body of the text. A “virtual disc” with a simplified geometry was added to the LAA to represent the properties of the more complex physical geometry of the AMPLATZER™ Amulet™ device.

The computational model was based on applying a force to the center of the disc (equivalent to the tug force exerted by the operator) and solved to find the circumferential resultant force at the boundary between the disc and the tissue, as well as the disc displacement. To simplify the model, fixed entrances and exits to the left atrium (the pulmonary veins and the mitral valve) were assumed. Structural contact between the circumference of the device lobe and the opening to the LAA was established by ensuring continuity between these elements. These conditions represent the assumptions used in the simulation.

By inputting the measured elastic properties (stress-strain data) of the device and assuming elastic behavior of the tissue, the resultant force and related stress were determined through previously established equations [8].

The stresses were integrated over the contact area of the device with the LAA tissue to yield cumulative forces experienced on the walls by device pulling (i.e., the applied force). The circumferential forces from this integration represented 46% of the value for forces applied via the computation at the virtual disc midpoint, assuming linearity.