Residual Leaks Post Left Atrial Appendage Occlusion
Reported from ACC 2022
Michele Pighi provides his take on the National Cardiovascular LAAO Registry that analyzed data from 51,333 patients who underwent LAAO procedures with the WATCHMAN device between 2016-2019, as part of the College's NCDR registries.
Why this study – the rationale/objective?
Percutaneous LAA occlusion (LAAO), using dedicated devices, has emerged as a safe and useful alternative to oral anticoagulation in those patients who present with contraindications, intolerance, high bleeding risk, refuse such therapy. However, the incidence of post-procedural leaks (of any size) ranges between 12.5% and 50.8% in major registries and randomized trials. To date, few data are available about the clinical impact of leakage to the left atrial appendage due to incomplete device sealing after left atrial appendage occlusion (LAAO), suggesting no association with thrombotic events, but the number of events was very limited, and the studies underpowered to detect any difference. Conversely, recent data from. a combined analysis (PROTECT AF and PREVAIL and CONTINUED ACCESS studies) found that leaks at 1 year (but not at 45 days) are associated with higher rates of ischemic strokes.
How was it executed - the methodology?
The National Cardiovascular Data Registry LAAO Registry analyzed data from 51,333 patients who underwent LAAO procedures with the WATCHMAN device between 2016-2019, as part of the College's NCDR registries. Patients were classified according to leak size on echocardiography at 45 post-procedure [0 mm, no leak (73.4%); >0-5 mm, small leak (25.8%); and >5 mm, large leak (0.7%)]. Data were collected at 45-day, 6-month and 1 year. Registry data also included information about subsequent adverse health events occurring in the years after the procedure. The primary endpoint was a composite of stroke, TIA, or systemic embolization. The secondary endpoints were major bleeding, death, and major adverse events.
What is the main result?
The researchers found a modest but statistically significant difference in terms of baseline characteristics, with a higher incidence of non-paroxysmal AF, cardiomyopathy and larger LAA orifice diameters in patients presenting with large leaks. There were no differences in terms of moderate sedation or ICE usage, case duration, contrast volume, or in-hospital complications. The majority (>90%) of the patients were discharged on anticoagulation, with a decrease at 45 days, with a large proportion of patients presenting with larger leaks remaining on an anticoagulation regimen up to six months and one year.
In regard to the primary endpoint, patients presenting with small leaks compared with the no leak group showed a higher event rate at one year (HR 1.15 (95% CI 1.02-1.29). Because of the small dimension of the large cohort, no significant differences were detected compared to the other populations. However, the absolute incidence rates were small (<4%) in all three groups. The small leak group presented with a higher event rate of major bleedings HR 1.10 (95% CI 1.03-1.20) compared to patients without any leak, although no significant differences emerged in the rates of anticoagulant use at 1 year. All-cause mortality was numerically higher in patients with large leaks, but because of the small cohort, no significant differences were detected. Finally, the small leak group presented with a higher event rate of major adverse events (HR 1.10 (95% CI 1.05-1.16), compared to patients without any leak (most likely driven by the embolic and bleeding complications).
Secondary analysis focusing on any leak versus no leak populations confirmed a higher event rate for the primary endpoint (HR 1.15 (95% CI 1.02-1.29), while when stratifying for the dimensions, only leaks between 0-3mm reached statistical significance (HR 1.14 (95% CI 1.001-1.23) when compared to the no leak population.
Limitations
- Observational registry
- Imaging for leak >45 days was uncommon in clinical practice
- Variation in peri-device leak size measurement and reported
- Only the first-generation Watchman device included
- Data on interventional leak management were not available (e.g., coils and plugging devices)
- Follow-up limited to 1 year
Conclusion and interpretation
This is the largest study to date reporting on the clinical significance of leaks following LAAO, showing that peri-device leak at 45-day follow-up is observed in around 25% of patients following LAAO with Watchman 2.5 device, with the majority of the leaks being ≤5 mm (>99% patients). Finally, the presence of peri-device leaks <5 mm was associated with a ~10-15% increase in 1-year risk-adjusted rates of thromboembolic and bleeding complications.
Of importance, the presented data reflect prior practice using an older device, while the novel iteration Watchman Flex showed a ~10% rate of residual leaks. On top, operators are progressively gaining more experience and access to planning tools which may help in refining the device sizing and implantation. Finally, in regard to the possible interventional approach to small leaks, very few data are available (case reports/series only), warranting further studies to investigate the best management strategies.
No comments yet!