1 INTRODUCTION

The left atrial appendage (LAA) is a highly variable blind-end cystic structure composed of pectinate muscle (comb muscle) and trabecula. An increasing number of studies have demonstrated that the morphology of the LAA is closely associated with the risk of stroke, with more than 90% of thrombosis occurring in the LAA among patients with nonvalvular atrial fibrillation (NVAF).^{1, 2} Currently, left atrial appendage occlusion (LAAO) is considered an ideal alternative treatment for preventing embolic stroke in patients with NVAF.³ Standard LAAO (SLAAO) typically involves the use of X-rays, general anesthesia, endotracheal intubation, mechanical ventilation and transesophageal echocardiographic (TEE) guidance, especially in those centers at the beginning of their experience. In some cases, however, SLAAO may increase the risks associated with anesthesia and esophageal injury, leading to delayed patient recovery, prolonged procedure time, increased costs, and extended hospitalization. In addition, SLAAO guided by TEE is not feasible for patients with gastrointestinal or stomatological diseases. If fluoroscopic assessment can replace TEE guidance, it would significantly simplify the procedure, reduce complications related to TEE, intubation, general anesthesia, lower operation costs, and shorten both the procedure and postoperative recovery time. Emerging evidence suggests that LAAO guided by fluoroscopy guidance only (minimal LAAO [MLAAO]) is also safe and reliable.^4-7 The aim of this study was to summarize the preliminary results of MLAAO in patients with NVAF under the guidance of fluoroscopy at our center, assess the feasibility and safety of MLAAO and explore its potential benefits.

2 METHODS

A total of 254 consecutive patients diagnosed with NVAF who were scheduled to undergo LAAO using the WATCHMAN™ device (Boston Scientific) at Taizhou People's Hospital affiliated with Nanjing Medical University from January 2019 to October 2021 were initially included in the study. Five patients were excluded: three due to the presence of LAA thrombus, and two due to an excessively large LAA. Consequently, a total of 249 patients were retrospectively analyzed, consisting of 148 males and 101 females, aged between 58 and 86 years, with an average age of 69.61 ± 5.23 years. The selection of either the SLAAO or MLAAO approach was based on informed consents made by the patients. Ultimately, 183 patients were assigned to the SLAAO group, while 66 patients were assigned to the MLAAO group. All patients met the following inclusion criteria: age > 18 years, NVAF, high risk for stroke (CHA2DS2-VASC score ≥ 2 for males and ≥3 for females), intolerance to long-term oral anticoagulant therapy, contraindications to anticoagulation, or oral anticoagulant therapy failure (e.g., stroke or transient ischemic attack during anticoagulant therapy). The study was approved by the Ethics Committee of Taizhou People's Hospital affiliated with Nanjing Medical University, and written informed consent was obtained from all patients before the procedure. The procedure was performed by two physicians with a minimum of 40 prior LAAO procedure volumes. Successful implantation was defined as stable device positioning with no or residual peri-device leak (PDL) width <5 mm. Furthermore, the analysis of digital subtraction angiography (DSA) data during the procedure in both groups was performed by a specialized engineer using Siemens DSA software.

2.1 CT angiography (CTA) examination for preprocedural and follow-up imaging

CTA with dual-source CT (Siemens SOMATOM force) was performed 24 h before LAAO to exclude the presence of thrombus in the left atrium (LA) or LAA, and to measure the ostial dimensions and depths of the LAA. Additionally, routine CTA was performed at 3 months postprocedure for all patients. In brief, 80 mL of nonionic contrast agent (Iomeron 350) was injected at a rate of 4−5 mL/s. The scanning parameters were as follows: voltage 100-120 kV, effective tube current 800−1235 Ma, collimator width 256 × 0.625 mm, frame rotation time 270 ms, and spacing 0.2. A two-phase scan was performed, with the second scan delayed by 60 s from the first phase scan. Thrombosis was ruled out if no low-density filling defects were observed during the dual-phase scanning. The presence of blood stasis without thrombosis was defined as the appearance of a filling defect in the first phase that disappeared in the second phase after the 60-second delay. The presence of thrombus was defined as a low-density filling defect observed in the first phase, and persisted in the second phase. When the scan was completed, the data were transferred to the Syngo VB10 workstation (Siemens), and image analysis was performed using MM reading software, including three-dimensional reconstruction (3DR), volume rendering (VR), and maximum intensity projection (MIP), to obtain multi-directional anatomical information of the LAA and Watchman occluder images (Figure 1).

2.3 MLAAO procedure using the watchman device

In regard to MLAAO, the procedure was performed with local infiltration anesthesia of 1% lidocaine. The transseptal puncture site was determined as inferior−posterior based on the fluorescence image of the anterior−posterior position and right anterior oblique (RAO) 45°. Following a successful transseptal puncture, contrast angiography was performed in three positions, RAO 30°/caudal (CAU) 20°, RAO 30°, and RAO 30°/cranial (CRA) 20°, using a sheath and pigtail catheter to visualize the morphology of the LAA. The largest LAA orifice diameter, measured with calibration of the marker on the delivery sheath in the three views, was considered the final diameter of the landing zone. Based on the measured LAA diameter from CTA and DSA, a Watchman device with an oversized diameter of 4–6 mm was selected and deployed under fluoroscopic guidance. It should be noted that a previous study indicated that the ostium diameter and depth of the LAA measured by CTA were larger than those measured by DSA [8]. In case of discrepancy between DSA and CT scan results, a larger-sized occluder could be considered on the basis of the DSA measurement results and the principle of 4-6 mm degree of device over-sizing. The “PASS” principle of fluoroscopy was then employed to assess (Figure 2): “P” (Position)—a tangential CAU 20°/RAO 0°−45° view was used to assess the exposure of the occluder's shoulder, ensuring it was located at the LAA ostium or slightly away (over-shoulder distance within 1/4–1/3 of the diameter of the Watchman device); “A” (Anchor)—stable anchoring was confirmed by a tug test without displacement under fluoroscopy; “S” (Size)—the occluder compression ratio was determined at the widest cross-section using tangent fluoroscopic views from three positions: RAO 30°/CAU 20° (equal to TEE 135°), RAO 30° (equal to TEE 90°), and RAO 30°/CRA 20° (equal to TEE 135°), and calibrated with the marker on the delivery sheath. The compression rate was then calculated as: (manufacturer device diameter—measured diameter by DSA)/manufacturer device diameter * 100%. A value between 8% and 25% was deemed acceptable; “S” (Seal)—the degree of sealing was evaluated by examining PDL in three fluoroscopic views: RAO 30°/CAU 20°, RAO 30°, and RAO 30°/CRA 20°. A result with no leakage or residual leakage < 5 mm were considered acceptable. If the “PASS” principle was satisfied, the Watchman device was then released. Following the procedure, transthoracic echocardiography (TTE) was performed to detect the presence of pericardial effusion.

3 RESULTS

3.2 Comparison of procedural characteristics between the SLAAO group and the MLAAO group

As shown in Table 2, there was no statistically significant difference in the rate of successful device implantation between the SLAAO group and the MLAAO group (98.91% vs. 98.485, p > 0.05). Furthermore, no statistically significant differences were observed in device resizing, deployment attempts per procedure, device size, the incidence of residual flow, and occluder shoulder exposure between the two groups (p > 0.05, respectively). However, the MLAAO group exhibited higher fluoroscopy time, radiation exposure dose, and contrast medium compared to those in the SLAAO group (p < 0.001, respectively). Moreover, the MLAAO group demonstrated significantly shorter procedure time and hospitalization duration compared to the SLAAO group (p < 0.001, respectively). Regarding the occluder compression ratio, a t-test was utilized to compare the values measured by TEE and fluoroscopy, and the correlation between TEE and fluoroscopy measurements were analyzed using a Spearman analysis in the SLAAO group. As presented in Table 2 and Figure 3, the occluder compression ratio measured with fluoroscopy was lower than that measured with TEE (t = 8.61, p < 0.001), but a good correlation (R² = 0.834) was observed. Consequently, there was no significant difference in the occluder compression ratio measured with fluoroscopy between the SLAAO group and the MLAAO group. However, when comparing the occluder compression ratio measured with fluoroscopy in the MLAAO group to that measured with TEE in the SLAAO group, the MLAAO group exhibited a lower rate (17.63 ± 3.75% vs. 21.69 ± 4.26%, p < 0.001). Significant differences were observed between the SLAAO group and the MLAAO group in terms of oropharyngeal/esophageal injury, hypotension, and dysphagia (p < 0.05).

Table 2. Comparison of procedural characteristics between the SLAAO group and the MLAAO group.

Procedural parameters	SLAAO group (n = 182)	MLAAO group (n = 67)	t/χ2 value	p Value
Implantation success rate, %	98.91	98.48	0.00	1.000
Device resizing, n (%)	3 (1.65)	1 (1.49)	0.00	1.000
Deployment attempts per procedure	1.14 ± 0.42	1.20 ± 0.48	0.96	0.169
Procedure time, min	57.51 ± 7.08	37.48 ± 6.89	19.95	0.000
Fluoroscopy time, min	10.64 ± 2.08	12.85 ± 2.12	7.39	0.000
Radiation exposure dose, mGy	617.52 ± 19.78	667.85 ± 23.29	16.95	0.000
Contrast medium, mL	41.96 ± 7.27	89.25 ± 6.19	47.29	0.000
Hospital stay, day	6.22 ± 0.49	5.14 ± 0.42	16.02	0.000
LAA morphology
Cauliflower, n (%)	116 (63.73)	41 (61.19)	0.14	0.708
Chicken wing, n (%)	29 (15.93)	12 (17.91)	0.23	0.632
Cactus, n (%)	24 (13.19)	11 (16.42)	0.42	0.517
Wind stock, n (%)	13 (7.14)	3 (4.48)	0.58	0.447
Device size
21 mm, n (%)	12 (6.59)	5 (7.46)	0.06	0.806
24 mm, n (%)	33 (18.13)	12 (17.91)	0.00	1.000
27 mm, n (%)	57 (31.32)	21 (31.34)	0.00	1.000
30 mm, n (%)	34 (18.68)	10 (14.93)	0.47	0.493
33 mm, n (%)	46 (25.27)	19 (28.36)	0.24	0.624
Residual flow
0 mm, n (%)	129 (70.88)	45 (67.16)	0.32	0.572
1−5 mm, n (%)	53 (29.12)	22 (32.84)	0.32	0.572
≥5 mm, n (%)	0 (0.00)	0 (0.00)	0	1.000
Exposure shoulder of occluders
Yes, n (%)	87 (47.80)	34 (50.75)	0.17	0.680
No, n (%)	95 (52.20)	33 (49.25)	0.17	0.680
Compression ratio of occluders
DSA, %	17.92 ± 4.09	17.63 ± 3.75	0.51	0.305
TEE, %	21.69 ± 4.26	—
Complications
Pericardial effusion, n (%)	2 (1.10)	1 (1.49)	0.06	0.807
Cardiac tamponade, n (%)	1 (0.55)	0 (0.00)	0.62	0.431
Vagal reflex, n (%)	0 (0.00)	1 (1.49)	1.10	0.294
Oropharyngeal/Esophageal injury, n (%)	10 (5.49)	0 (0.00)	2.11	0.146
Oropharyngeal hematoma, n (%)	3 (1.65)	0 (0.00)	1.12	0.289
Hypotension, n (%)	11 (6.04)	1 (1.49)	2.21	0.137
Dysphagia, n (%)	6 (3.30)	0 (0.00)	2.26	0.133
Fever, n (%)	3 (1.65)	1 (1.49)	0.01	0.920
Vascular complications, n (%)	2 (1.10)	1 (1.49)	0.06	0.807

• Abbreviations: MLAAO, minimal left atrial appendage occlusion; SLAAO, standard left atrial appendage occlusion.

4 DISCUSSIONS

The major findings of the present study are as follows: (1) Both SLAAO and MLAAO demonstrated high procedural success rates and safety. (2) Although the MLAAO group exhibited longer fluoroscopy time, higher radiation exposure dose, and greater contrast medium compared to the SLAAO group, there were significantly reductions in procedure time and hospital stay duration in the MLAAO group. Significant differences were also observed in terms of oropharyngeal/esophageal injury, hypotension, and dysphagia between the SLAAO group and the MLAAO group. (3) The occluder compression ratio measured by fluoroscopy was founded to be lower than that measured by TEE, but a good correlation was observed between the two methods. (4) Both SLAAO and MLAAO procedures demonstrated effectiveness in preventing embolic stroke in patients with NVAF, as evidenced by the 3-month follow-up results.

The SLAAO procedure typically involves general anesthesia, and is guided by CTA, TEE, and fluoroscopy. However, anesthesia-related risks are inherent in the SLAAO, necessitating the simplification of the LAAO procedure for the benefit of patients. The concept of such simplification has been explored for over 20 years, primarily focusing on Amplatzer occluders rather than Watchman occluders.^{9, 10} Currently, many centers utilize TEE guidance with sedation alone.^{11, 12} In some centers and cases, LAAO has even been performed without prior TEE or CT.¹³ TEE provides the advantage of confirming the absence of intracardiac thrombus before the procedure, evaluating the LAA size, shape, and atrial septum characteristics during puncturing, and detecting pericardial effusion. However, TEE is an invasive examination method requiring anesthesia and sedation, and carries potential risks associated with anesthesia and esophageal injury. Moreover, it is contraindicated in certain patients. Intracardiac echocardiography (ICE)-guided LAAO under local anesthesia has been considered a safe and effective alternative to TEE for percutaneous LAAO.^{9, 14, 15} However, its lower cost-effectiveness restricts its widespread clinical use. Therefore, the introduction of the MLAAO may provide an alternative procedure for a larger number of eligible patients requiring LAAO.

Regarding LAAO, CTA plays a crucial role as a visualized and measured scan. It serves as an important tool for thrombus elimination, demonstrating a weighted mean sensitivity and a negative predictive value of 100%.¹⁶ CTA measurements of the ostium diameter and depth of the LAA have been found to be greater than those obtained by TEE and DSA, with CTA showing the strongest relevance and concordance.^{8, 17} In our study, all patients underwent CTA 24 h before LAAO, with a 60-second delay. 3DR, VR, and MIP techniques were employed to obtain multi-directional anatomical information about the LAA and facilitate the selection of an appropriate occluder size. Our results demonstrated a high degree of consistency between the LAA morphology obtained from MIP and fluoroscopic imaging (Figure 4), underscoring the valuable guidance provided by CTA in LAAO,^{8, 18, 19} particularly in the context of MLAAO.

The safety and feasibility of MLAAO have previously been evaluated.^4-7 In our study, we found no statistically significant difference in the rate of successful device implantation between the MLAAO group and the SLAAO group. Similarly, there was no significant difference in device resizing, deployment attempts per procedure, device size, incidence of residual flow, and occluder shoulder exposure between the two groups. These findings are consistent with the previous reports.

It is important to note that the MLAAO group exhibited higher radiation exposure dose and contrast medium usage compared to the SLAAO group. This can be attributed to the need for multi-angle angiography to confirm the fluoroscopic “PASS” principle. It should be considered that patients with renal dysfunction may not be suitable candidates for MLAAO due to the increased contrast medium usage. However, we observed significantly shorter procedure times and hospital stay durations in the MLAAO group compared to the SLAAO group. Furthermore, there were significant differences in anesthesia and TEE-related complications, including oropharyngeal/esophageal injury, hypotension, and dysphagia between the two groups. These findings suggest that the MLAAO offers greater cost-effectiveness and safety compared to the SLAAO.

According to the “PASS” principle, the appropriate compression ratio of the Watchman occluder is an important factor in ensuring safety and preventing occluder displacement.²⁰ In SLAAO, the occluder compression ratio is often assessed using TEE. However, in MLAAO, the occluder compression ratio is assessed solely through fluoroscopy. Therefore, it is necessary to establish the consistency and relationship between the occluder compression ratio measured by TEE and fluoroscopy. In our study, we initially evaluated the consistency of the two measurement methods in the SLAAO group. The results revealed that the occluder compression ratio measured with fluoroscopy was lower than that measured with TEE, but there was a good correlation between the two measurements, indicating the reliability of fluoroscopy in assessing the occluder compression ratio. These findings are consistent with the reports by Zhang et al.⁴

Furthermore, we compared the results of the follow-up 3 months after LAAO between the SLAAO group and the MLAAO group using CTA. CTA serves as a suitable alternative to TEE for post-LAAO device surveillance and is more sensitive than TEE in detecting residual flow.²¹ Our analysis showed no significant differences in the incidence of device displacement, residual flow, DRT, stroke, or bleeding between the two groups. However, the incidence of residual flow within the range of 1–5 mm was higher in the MLAAO group compared to the SLAAO group. Additionally, one case in the MLAAO group exhibited residual flow exceeding 5 mm. The follow-up results confirmed the safety and effectiveness of MLAAO in preventing embolic stroke in patients with NVAF. However, they also indicated the presence of technical challenges and the need for optimizing MLAAO to achieve greater occlusion.²²

5 CONCLUSION

MLAAO, guided by fluoroscopy without the need for TEE and general anesthesia, appears to be safe, effective, and feasible, particularly for patients who are unable to tolerate or unwilling to undergo TEE and general anesthesia. Furthermore, MLAAO simplifies the procedural steps and facilitates the application of LAAO. However, it is important to note that further validation in a larger number of patients is needed, and there remain technical challenges in the optimization of MLAAO to achieve greater occlusion.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.