Discussion
To the best of our knowledge, this is the largest study of LAAC using the LAmbre device. We found a relatively high incidence of delayed PE/PT after LAAC with the LAmbre device using a conventional implantation method, mainly in patients with combined treatment. The MM for implantation of the LAmbre device significantly facilitated full opening of the umbrella, reduced the requirement of the recapture manoeuver, and decreased the incidence of PE/PT. In multivariate Cox regression analysis, an umbrella that was not fully open was the only factor independently associated with delayed PE/PT events.
PE/PT is a serious complication, which mostly occurs during LAAC, but rarely occurs after this procedure17. Schmidt et al18 found that the PE/PT rate was significantly higher in patients who were implanted with the Watchman device in the EWOLUTION study during sinus rhythm than in AF rhythm (0.2% vs 1.5%; P=0.02) at day 30 post-implantation. Delayed PT events that required intervention only occurred in LAAC during sinus rhythm. The investigators speculated that the main mechanism for their finding was LAA contraction during sinus rhythm exerting mechanical force on the device, which eventually led to PE/PT. Wolfrum et al19compared the efficacy and safety of LAAC with the Amplatzer Cardiac Plug according to the different position of the device disc. No evidence for a difference in the occurrence PE/PT was found between patients with complete versus incomplete Amplatzer Cardiac Plug disc coverage of the LAA ostium. In contrast, a recent study16 on the Amulet device reported a tendency toward more PE/PT events after LAAC with incomplete coverage of the LAA ostium by the proximal Amulet disc (P=0.07). This previous study suggested that suboptimal positioning of the device might be associated with worse periprocedural outcomes independent of repositioning or exchange of the device during the procedure. In our study, delayed PE/PT events were neither associated with the coverage position nor recapture operation as analysed by univariate and subsequent multivariate Cox regression analyses.
Huang et al10 reported encouraging clinical outcomes in patients with NVAF who had implantation of the LAmbre LAA occluder. Three PT events (timing of the events was not specified) were documented in their study. Park et al11 reported 2/60 (3.3%) patients who underwent LAAC with the LAmbre device and suffered from PT at days 8 and 33 after the procedure. Pericardiocentesis was required for these two patients. A recent study with a small sample size13 showed that 5 (29.4%) delayed PE events occurred in 17 patients who underwent LAAC with the LAmbre device during 3 months of follow-up. Data from LAAC with the LAmbre device are relatively inadequate and the safety of combining LAAC with the LAmbre device and catheter ablation is unknown. In the present study, five of six patients with delayed PE/PT required intervention in those with the combined procedure. Only one PE event occurred among patients with non-PAF who underwent isolated LAAC.
The umbrella of the LAmbre device has double hooks (eight small distal hooks engaging into the LAA wall and eight U-shaped large hooks trapping in LAA trabeculations) with the purpose of decreasing the risk of device embolization. However, these eight U-shaped large hooks are bare and slender. A narrow space with widespread trabeculations in the LAA frequently interferes with opening of the umbrella. A stronger mechanical force against the LAA wall might develop when any of these hooks are not fully opened. Therefore, full opening of the umbrella is important and is considered as one of the COST criteria (including umbrella deployed beyond the Circumflex artery; umbrella fully Open; peri-device with optimal Sealing (leakage ≤3 mm) for release of the LAmbre device; and device stability confirmed by the Tug test) for release of the LAmbre device. The modified implantation method in which the umbrella is initially deployed half open outside of the LAA and complete deployment into the landing zone facilitates full opening of the umbrella. In our study, all of the patients with delayed PE/PT showed incomplete opening of the umbrella in fluoroscopic imaging in the CM group. Additionally, incomplete opening of the umbrella was significantly associated with occurrence of delayed PE/PT events as shown in the univariate model and multivariate Cox regression analysis. The nadir of the U-shaped hooks at the same level as the radio-opaque marker connecting the umbrella and the cover as shown in fluoroscopic imaging is an important sign of full opening of the umbrella. However, this is still insufficient to guarantee an appropriate mechanical force from the double hooks against the LAA wall because full opening of the umbrella was found in another two patients with delayed PT in the MM group. One of these patients required surgical management because a small branch of the left circumflex artery was likely injured by the small hooks. This case was considered as an exception. In patients with the combined procedure, contractile function of the LAA was gradually restored when sinus rhythm was maintained after the procedure. We speculate that relative movement between the bare large hooks and the LAA wall causing friction during sinus rhythm was the leading cause of perforation. Therefore, we strongly recommend using this MM for patients with the combined procedure or isolated LAAC in those with PAF. Performance of routine echocardiographic follow-up in this patient population may be necessary. The manufacturer of the LAmbre device might manage with the eight U-shaped big hooks in the next-generation product design.