Measurement of LAA Dimensions
The LAA orifice diameter were measured by both angiography and TEE. The mean diameter of LAA orifice measured by angiography and TEE was 19.0±2.9 mm and 16.1±2.0 mm (P<0.05), respectively; and the correlation coefficient was 0.65. When the angiographic and TEE measurements were not in agreement, the angiographic measurement was used. The mean distance from MA to LSPV measured by TEE was 26.4±2.8 mm.
Acute procedural outcome
The LAAs in 27 dogs were successfully occluded by the first chosen device of the normal type, while in 2 cases, it was replaced by the plus type. The mean diameter of the LAMax’s cover-disc in vitro before LAAC was 25.8±2.5 mm. After LAAC, the mean diameters of cover-disc measured by TEE and angiography were 23.8±2.6 mm and 24.0±2.8 mm (P>0.05), respectively, and the correlation coefficient was 0.97, and both of them were less than that of the LAMax’s cover-disc in vitro.
Follow-up by TEE
Scheduled TEE examinations revealed that the LAAC devices were well localized, no migration and thrombus on the cover-disc were observed. At the immediate post-LAAC, there were 5 cases of residual flow detected by color Doppler flow imaging (CDFI): 1 case with a flow width of 5 mm, and the others lower than 3 mm. Yet, the residual flow in three cases was not detected at the follow-up visit. TEE revealed that the cover-disc in 18 dogs (62.1%) was completely pulled into the LAA without residual flow and formed concavity, and in 11 dogs (37.9%) incompletely pulled into the LAA formed suboptimally concavity, while 5 of them (45.5%) had residual flow (Table 1). Additionally, TEE examinations revealed 3 cases of MA compression, which was reduced in 2 out of 3 cases at a 1-month follow-up and did not affect the movement of the posterior valve (Table 1).
Anatomical Examination
Anatomical examinations showed that all devices were placed within LAA, the LAA ostia were well occluded. In the dogs euthanized at 1-, 2-, 3-, and 6-months after follow-up, the left atrial surface of the discs was covered by a glistening white pannus layer (Figures 2-4), and no thrombus was found. No visible infarcts were detected in the major organs.
Healing response of cover-discs
The healing was faster in the dogs whose cover-disc were completely pulled into the LAA, and the neo-intima was well covered on the atrial surface of the disc and connected with surrounding tissues; therefore, the LAA was completely blocked by LAMax device. As shown in Figure 2A, both the atrial surface and central screw hub of cover-disc was well covered by neo-intima at 1-month follow up. Yet, the healing was delayed in dogs whose cover-disc were incompletely pulled into the LAA with suboptimally concavity (Figure 2B).
If a cover-disc which was 8 mm larger than the measured LAA orifice was incompletely pulled into the LAA, the surrounding structures would have been affected by the disc. The granulomas were found at the part of MA that was in contact with the cover-disc in one dog, and the flow of LSPV was partially blocked, but the mitral valve movement was not affected (Figure 2C). In this case, the anchor was placed in a small lobe of the bi-lobular LAA, and the cover-disc was not pulled into the LAA, the part of MA that was in contact with the disc also found the granulomas (Figure D, E).
The residual flow detected by TEE was found related to the cover-disc, which was incompletely pulled into LAA. As shown in Figure 2F, with reference to the residual flow, the surface of the cover-disc was completely covered by neo-intima. However, the part at 180o did not connect with LA wall, and there was an irregular fissure.
The completed coverage of endothelialization on the central screw hub of cover-disc was found in 2 cases in the groups undergoing follow up at 1-month, 2 at 2-months, 2 at 3-months, and 2 at 6-months after LAAC.
Healing response of anchor part
TEE revealed that the cover-disc was not attached to the LAA ostium in 2 out of 29 dogs with a certain distance, and the anchor was localized at the LAA neck. A thin layer of neo-intima was found on the surface of the cover-disc and anchor; and partially on the short central waist in one dog after 1-month of LAAC (Figure 3 A, B). After 3-month of LAAC, the surface of the cover-disc and the anchor were completely covered by neo-intima, and the short central waist was completely covered by neo-intima (Figure 3 C, D). Both of them had the immediate residual flow after LAAC, but disappeared because of the endothelialization of the anchor.
LAA cavity closure
In all cases, the anchors were well localized in the LAA, and the claws of anchors were not pierced through the LAA wall. The short central waist was not broken in any case. Since the distance between the cover-disc and the anchor was short, the walls of LAA in the dogs without residual flow after 2-months of LAAC could not be separated. However, the LAA cavity of the dog with the residual flow was still present since flow communication disturbed and delayed tissue healing.
Histological examination
At 1-, 2-, 3-, or 6 months follow-up, the atrial surfaces in all cover-plates of the LAMax device were covered by neo-intimal layers. The histological sections showed that a moderate granulomatous inflammation was found near the woven material, the neo-intima spread into the LA wall surface, and across the device-left atrial interfaces to completely seal the LAA openings, which were covered by a thin layer of endothelial cells (Figure 3). Endocardium, which consisted of smooth muscle cells in a proteoglycan-collagenous matrix, was found near metal and fabric. The claws of the anchoring parts were well opposed to the LAA walls, with no evidence of tissue necrosis. Retention hooks were seen embedded in the LAA walls. No infarcts were observed in the major organs.