Comparison of Left Atrial Appendage Occlusion vs Standard Medical Therapy in Patients in AF Undergoing TAVI

Background and Rationale

Transcatheter aortic valve implantation (TAVI) is a minimally invasive treatment option for patients suffering from severe symptomatic aortic stenosis. Superiority of TAVI over medical treatment has been shown in the randomized PARTNER B trial followed by the PARTNER A trial showing non-inferiority over a more invasive open-heart surgical aortic valve replacement in high-risk patients. Over the last years, many patients have been successfully treated with TAVI worldwide.

Peri- and post-procedural morbidity differed considerably between the surgical and interventional treatment group in the PARTNER A trial: while major bleeding (19.5% vs 9.3% at 30 days) and new onset atrial fibrillation (16% vs 8.6% at 30 days) occur more often in surgically treated patients, vascular access complications (3.8 vs 17% at 30 days) and neurologic events (5.5% vs 2.4% at 30 days) are more frequently encountered in the TAVI population. The higher rate of neurologic events in the TAVI group was of particular concern, given its association with a higher mortality and the clinical impact of major neurologic events on patient's quality of life and daily functioning. Only about 40-50% of strokes occur intra-procedurally, followed by a higher hazard rate in the first week and a constant hazard thereafter. After 30 months, the surgical group experienced a numerically higher stroke rate, although not statistically significant. Measures to reduce procedural and post-procedural stroke rate were and are still looked for.

While procedural strokes can be most likely reduced by improvements in valve delivery systems (smaller delivery systems) and more operator experience, cerebral protection devices are currently tested in trials but preliminary results are rather disappointing. Risk factors for post-procedural strokes after TAVI are a previous history of stroke, more extensive peripheral vascular disease, and higher functional class.

A subgroup of patients of particular concern are patients suffering from atrial fibrillation: predictor of early (day 1-30) cerebrovascular events was new-onset atrial fibrillation (OR 2.76), whereas chronic atrial fibrillation was a predictor (OR 2.84) for the occurrence of late (>30 days) cerebrovascular events. Amat-Santos reported a higher rate of strokes/systemic embolism in patients with in hospital new onset atrial fibrillation after TAVI (13.6% vs 3.2% after 30 days). Patients diagnosed with new onset atrial fibrillation during or after the procedure in which no anticoagulation was initiated experienced an alarmingly high 30-day stroke rate of 40%.

Despite these results and the high prevalence of atrial fibrillation (AF) of about 33% in the overall TAVI population, only 30% of AF patients were on vitamin K antagonists (VKA) before, and only 70% after TAVI. While oral anticoagulation is an effective medical treatment for ischemic stroke prevention in these patients, it is often withheld from patients given their high risk of bleeding. TAVI patients typically not only suffer from much co-morbidity and are poly-medicated, but also suffer from a lot of bleeding risk factors. The bleeding risk score (HASBLED score) expresses the risk of major bleeding with a scoring system consisting of 9 individual risk factors for bleeding; a patient is at high risk for bleeding, if 3 or more factors are present. The stroke risk score (CHA2DS2-Vasc score) comprises of 9 individual risk scores for stroke in patients suffering from atrial fibrillation. Both risk scores share common risk factors, such as age, hypertension and previous stroke. It is therefore not surprising that patients at highest risk for stroke are at the same time at highest risk for major bleedings. The TAVI population forms such a high risk group with a high prevalence of the above risk factors: average patients age in the PARTNER trial was 84 years, with 11% showing severe renal impairment, about 75% had an indication for concomitant acetylsalicylic acid use and all suffered per definition from heart failure. In Stortecky's study on TAVI patients suffering AF, the average CHA2DS2-Vasc score was 4.5, with >95% of patients having a score >3.

Left atrial appendage occlusion (LAAO) allows avoiding oral anticoagulation and provides at the same time an at least equally good protection from strokes and peripheral embolism. It may therefore be an attractive alternative to oral anticoagulation in the TAVI patient population: the concept of LAAO is based on the fact that thrombus formation in atrial fibrillation occurs in >90% in the left atrial appendage (LAA). Mechanical occlusion of the LAA reduces the stroke risk by eliminating the source of thrombus formation.

Clinical studies proved this concept to be true: the PROTECT-AF study randomized >700 patients to either LAA occlusion or medical therapy (VKA). At one year, non-inferiority of LAAO was proven, with numerically less embolic events in the LAAO group. At a European conference (EuroPCR conference) 2013, Holmes (principal investigator of the PROTECT-AF trial) reported 4-year follow-up data with a significant 40% reduction of the composite of stroke/peripheral embolism/cardiovascular death and a 34% reduction of all cause mortality in the LAAO group as compared to oral anticoagulation. By reducing bleeding complications, it can be anticipated that morbidity and mortality will continue to diverge between the two groups. Comparable results from a large registry using Amplatzer devices (St. Jude Medical, St. Paul, USA) for LAAO were reported: after an average follow-up of 32 months, cardiovascular death, stroke, and peripheral embolism occurred in 7% of patients. In the latter study, oral anticoagulation was instantly stopped after LAAO in all patients.

In the here proposed "Randomized Comparison of Left Atrial Appendage Occlusion versus Standard Medical Therapy of Atrial Fibrillation in Patients Undergoing Transfemoral Transcatheter Aortic Valve Implantation", study the investigators test the hypothesis, that LAAO is superior to standard medical therapy in the high-risk TAVI population. This hypothesis has not been investigated by previous studies so far. The design is an investigator-initiated, randomized, multicenter, non-blinded, all-comers study. It is understood, that this is a pilot trial with the intention to prove safety, but which is underpowered to show superiority over medical therapy.