Antithrombotic therapy post TAVR: A double edged sword? ESC 2018: Insights from the French TAVI registry

A multicenter, prospective nation-wide French registry evaluated whether oral anticoagulation therapy was an independent correlate of long-term survival and early bioprosthetic valve dysfunction (BVD) in patients who underwent successful Transcatheter Valve Implantation (TAVI). The French registry was launched in 2007 and involved 11,469 patients with a mean duration follow-up was 495±3.5 days. This registry contained 11 years of data and analyzed patients from January 1, 2013, and December 31, 2015.

The aim was to assess independent correlates of long-term mortality and early bioprosthetic valve dysfunction (defined as increased prosthetic gradient ≥10 mmHg or new gradient≥20 mmHg) and to investigate whether anti-thrombotic treatment influenced long-term mortality and early BVD after successful TAVI according to treatment at discharge, whether the patient was on anticoagulation or not. The hypothesis behind this was that anticoagulation may reduce bioprosthetic and non-bioprosthetic driven events.

The investigators had the full survival data on 90% of the patients. Data was not available for those without known anticoagulation on discharge or those who died during the hospital stay. Roughly 40% of patients were discharged on anticoagulation. The investigators evaluated baseline characteristics and looked at independent correlates associated with the use of anticoagulants. The most common cause of anticoagulation was atrial fibrillation, with 70 percent of patients on OAC due to atrial fibrillation. Neither aspirin nor clopidogrel was independently associated with mortality. There was also some difference in mortality in terms of procedures which were non-femoral access TAVR ( HR=1.18 [1.04-1.35], p=0.011). Interestingly, male gender (adj HR 1.63 [1.44-1.84], p<0.001), chronic renal failure (adj. HR 1.37 [1.23-1.53], p<0.001),  AFib (adj. HR 1.41 [1.23-1.62], p<0.001) and anticoagulation at discharge (adj. HR 1.18 [1.04-1.35], p=0.013) were all associated with mortality.

On the contrary, the results for bioprosthetic valve dysfunction mainly related to valve thrombosis showed that anticoagulation was associated with less bioprosthetic valve dysfunction (adj. OR 0.54 [0.35-0.82], p=0.005).

These findings reflected the complex interactions and existence of multiple independent correlates of long-term mortality and valve thrombosis, following successful TAVI. The potential clinical benefit of oral anticoagulation was a matter of debate. On one hand, OAC was associated with a reduced risk of valve thrombosis.  However, it varied collinearly with atrial fibrillation and still remained independently associated with mortality, despite adjustment for atrial fibrillation. Emphasizing on the need for further research,  Dr. Jean Phillippe Collet, Head of the Action Study Group, Sorbonne Université in Paris stated, “The major question that remains unanswered is whether anticoagulation itself is detrimental or whether it is because of the indication.” To address this further and rule the presence of unknown confounders, Dr. Collet acknowledged the need for randomized evidence to aid in a more robust conclusion. Referring to antithrombotic therapy as a two-edged sword, Dr. Collet concluded, “The missing link is between bioprosthetic valve dysfunction and mortality or stroke, and this relationship is yet to be determined.”