Transcatheter aortic valve replacement and thrombus formation

Leaflet thrombosis (LT) following transcatheter aortic valve replacement (TAVR) has been an enigma without a true cause identified. Several working hypotheses surround the aetiology and are aimed at recipient coagulation profile changes systemically and locally within the new valve, TAVR valve preparation and preservation, TAVR valve implantation technique [1, 2]. The reader should remember that TAVR does not replace the aortic valve; it excludes the native aortic valve or a prior bioprosthetic valve as in valve-in-valve. TAVR occupies the aortic root by displacing and opening the native aortic leaflets, reducing the coaptation of the native aortic leaflets and reducing the effective volume of the sinuses of Valsalva. Another important factor that the reader should be aware of is that the position of the TAVR valve commissures does not align with native aortic root commissures anatomically which is not the case with surgical aortic valve replacement. While this may not be relevant, it should be still worth investigating experimentally to determine whether methods to align TAVR commissures with native valve produce favourable fluid dynamics. Finally, there is more total mass occupying the aortic root (TAVR plus native aortic leaflets) in patients with TAVR compared to surgical aortic valve replacement patients (valve device only). With TAVR in place, there are 2 sinuses and new fluid dynamics components; there are new interfaces between the nitinol TAVR frame and the native aortic leaflets, aortic annulus and aortic wall, especially with the larger profile self-expandable valves such as Evolut valves (Medtronic). Each of these aspects alone produces new vortices and turbulences that are in essence primary sources for sinus stasis: one of the pillars of Virchow’s triad (stasis, changes in blood constituents and changes in the vessel lining) [1–3]. In vitro studies of valve turbulence, stress, Reynolds number and platelet activation threshold have been shown to be meaningful in the aetiology of coagulation cascade activation [4]. Furthermore, an interesting in vitro mode using coagulated milk has shown similarities to computed tomography findings that most of the thrombus is to the sinus followed by the aortic wall followed by the commissures and leaflets [5]. This suggests that different reasons could be culprit for different anatomical thrombi within the aortic root. In this well-executed meta-analysis by D’Ascenzo et al. [6], a total of 18 studies on LT in TAVR were meta-analysed for incidence of stroke. The authors carefully evaluated the presence of LT and clinical symptoms over the period of time. Limited by the data in each study, 12 studies reported the incidence of LT according to the presence or absence of oral anticoagulation and 6 studies quantified the occurrence of stroke in patients with or without LT. Only 6 studies reported clinical thrombosis and 6 other studies focused on subclinical thrombosis. Median follow-up was 1 year. Median age was 82 years and median predicted risk of morality was 5.8%. Despite heterogeneity in the pathophysiological variables (diabetes, atrial fibrillation and severity score), pooled incidence of LT was 0.43% per month. LT was associated with a 4-fold increased risk of stroke. Independent predictors were large diameter of the TAVR valve, valve in valve, single antiplatelet therapy, body mass index above 30 kg/m2 and balloon expandable devices. In a mechanistic study using the Resolve Registry (Assessment of TRanscathetEr and Surgical Aortic BiOprosthetic Valve Thrombosis and Its TrEatment With Anticoagulation (RESOLVE)) thrombosis and its treatment with anticoagulation, overexpansion of the balloon expandable valve, and deeper implantation of the self-expanding valve with lower cardiac output were also associated with an increased thrombus burden [3]. While it comes as no surprise that space occupying hardware such as TAVR or surgical aortic valve replacement in the aortic functional unit alters the fluid dynamics and vortices and with that comes an increased risk for thrombus formation, there still remain(s) elusive recipient-device interaction(s) yet to be identified that contribute(s) to the risk of LT. Fortunately, LT, while carrying a definite stroke risk, can be easily treated with anticoagulation, which reduces the risk of stroke. It is such summary statistics in this meta-analysis that address and define the problem triggering simulation and engineering models to determine the mechanism of LT and offer possible solutions. Nonetheless, valve thrombosis has been associated with early valve failure and not just strokes alone. This meta-analysis provides one strategy identifying the pathophysiological and technical aspects that can be addressed, possibly modified and applied in clinical practice to reduce the risk of LT and its clinical sequelae.