Author + information
- Pablo Salinas, MD⁎ (, )
- Raul Moreno, MD,
- Angel Sanchez-Recalde, MD,
- Santiago Jimenez-Valero, MD,
- Guillermo Galeote, MD,
- Luis Calvo, MD,
- Ignacio Plaza, MD and
- Jose Lopez-Sendon, MD, PhD
- ↵⁎Interventional Cardiology, Department of Cardiology, University Hospital La Paz, Castellana 261, Madrid 28046, Spain
We read with interest the paper by Amat-Santos et al. (1) describing the importance of new-onset atrial fibrillation (NOAF) after transcatheter aortic valve implantation (TAVI). They describe an association between NOAF and embolic events (EE), suggesting that NOAF may be a mechanism for late neurological events after TAVI. However, we believe that the adjudication of the EE overestimates the association and impairs the generation of a hypothesis of causality.
The association of TAVI with NOAF and EE is biologically plausible following Bradford Hill's classic causality criteria (2). But another criterion is temporality: the cause must precede the effect. Nonetheless, Patient #4 and Patient #6 from Table 4 (1) had the embolic event 48 h and 18 days, respectively, before the NOAF episode (1). Therefore, those EE should not be attributed to NOAF. Patient #10 from Table 4 (1) had a stroke 645 days after TAVI (640 days after NOAF), and moreover, was under warfarin treatment with an international normalized ratio of 2.2 at hospital admission. Thus, it is doubtful that this stroke may have a causal relation to the TAVI procedure.
Besides the statistical limitation of a low number of events, if we remove these 3 events from the NOAF group, there are 3 remaining strokes at 30 days (incidence 6.8% vs. 3.2% in the no-NOAF group; odds ratio: 2.22, 95% confidence interval: 0.43 to 11.47, unadjusted p = 0.330). The incidence of stroke/systemic embolism at 30 days would be 9.1% versus 3.2% (odds ratio: 3.03, 95% confidence interval: 0.65 to 14.18, unadjusted p = 0.141). Thus, the incidence of EE at 30 days does not reach statistical significance, and has a poor clinical and causal significance. The cumulative incidence at follow-up would only add 1 event with uncertain causal relationship to TAVI.
Stroke is a major concern in TAVI, accounting for as many as 11% of deaths at 1 month (3). In a smaller series with 91% transfemoral TAVI, our group found 6% NOAF, with no impact on EE (6%) or mortality during a mean follow-up of 11 months (4). Left atrial appendage might be a source of thrombus in NOAF after TAVI, but is still unconfirmed. It would be revealing if the authors provided information about transesophageal echocardiography. The presence of thrombus in the left atrial appendage before or during the TAVI procedure could contribute to support the AF hypothesis as a mechanism of the late embolic events.
In conclusion, we believe that this comprehensive study contributes largely to the current knowledge on TAVI; nevertheless, it provides insufficient evidence to establish a causal association between NOAF and EE after TAVI. The true origins of stroke during or after a TAVI procedure are still obscure (and most likely multifactorial). Bradford Hill's conclusion was, “All scientific work is liable to be upset or modified by advancing knowledge. That does not confer upon us a freedom to ignore the knowledge we already have, or to postpone the action that it appears to demand at a given time” (2). Thus, we should optimize antithrombotic treatment in these patients, but it is essential to support the research in this field.
- American College of Cardiology Foundation