Stroke rates after transcatheter aortic valve replacement: does valve choice play a role?

This commentary refers to ‘Comparison of newer generation self-expandable vs. balloon-expandable valves in transcatheter aortic valve implantation: the randomized SOLVE-TAVI trial’, by H. Thiele et al., 2020;41:1890–1899.

We agree that cerebrovascular accidents are among the most serious adverse events in patients undergoing transcatheter aortic valve implantation (TAVI). Accordingly, the lower stroke rate in the self-expanding valve (SEV) arm of the SOLVE-TAVI trial is an interesting finding.¹ However, the trial was only powered for equivalence of the primary combined endpoint and any observed difference in stroke rates between the SEV in comparison to the balloon-expandable valve (BEV) group is only hypotheses-generating and requires validation in larger randomized clinical trials.

Evidence from other randomized trials comparing SEV vs. BEV is limited but consistently showed numerically lower stroke rates with SEV. In the CHOICE trial, 30-day stroke rates were 2.6% for SEV vs. 5.8% for BEV (P = 0.33).² Interestingly, clinical valve thrombosis occurred in seven BEV patients (7.3%) and one SEV patient (0.8%; P = 0.06) at longer follow-up, which may also possibly explain lower stroke rates. In the SCOPE 1-trial, stroke rates were 2.0% in the SEV and 3.0% in the BEV group (P = 0.33).³

Evidence from large-scale registries with propensity matching such as the FRANCE-TAVI registry showed comparable rates of stroke with SEV (2.4%) vs. BEV (1.8%).⁴ In contrast, in the CENTER registry, stroke occurred less frequently in BEV (1.9%) vs. SEV (2.6%; P = 0.03).

Mechanisms of stroke are of great interest to implement measures for prevention. As such, the timing of stroke occurrence plays an important role. In SOLVE-TAVI, stroke occurred <24 h in the one patient in the SEV group whereas 7 out of 10 patients had stroke <24 h in the BEV group. Sub-acute stroke within <30 days occurred in three BEV patients. Post-dilatation may be related to higher stroke rates. However, it is questionable if this mechanism was a prominent factor in the SOLVE-TAVI trial, as post-dilatation was performed more frequently in the SEV compared with the BEV arm (35.2% vs. 22.9%, P = 0.006). In comparison to the CHOICE trial, post-dilatation rates were much lower with current generation SEV devices. Baseline atrial fibrillation may also be associated with periprocedural and long-term stroke rates. In SOLVE-TAVI, baseline atrial fibrillation rates were higher in comparison to other trials and observational studies reflecting the high-risk population. Strict adherence to oral anticoagulation may have prevented sub-acute and later cerebro-embolic events not only due to atrial fibrillation but also due to prevention of valve thrombosis. Taken together, prevention of stroke after TAVR is of major importance but needs much more data derived mainly from adequately powered randomized clinical trials.⁵

Conflict of interest: none declared.

References