The hospital results and 1-year outcomes of transcatheter aortic valve-in-valve procedures and transcatheter aortic valve implantations in the native valves: the results from the Swiss-TAVI Registry.

Abstract

The SwissTAVI Registry includes all consecutive patients undergoing transcatheter aortic valve implantation (TAVI) and valve-in-valve (VinV) procedures for a failed bioprosthesis in Switzerland. We report the real world, all-comers, 30-day and 1-year outcomes of patients undergoing VinV and standard TAVI procedures. Prospectively collected data from the 2 groups (VinV and standard TAVI patients) were retrospectively analysed. In an adjusted analysis, in-hospital and 1-year outcomes of VinV patients were compared with those of patients undergoing TAVI for native aortic valve disease in the same registry. A subanalysis of VinV procedures in stenotic or regurgitant bioprosthesis was also performed. Between February 2011 and December 2016, 4599 and 157 consecutive patients underwent TAVI in native aortic valves and VinV procedures in degenerate bioprosthesis, respectively. VinV patients were younger (78 ± 9.1 years vs 82.2 ± 6.3 years; P < 0.001) but at a higher risk for surgery (the logistic EuroSCORE: 28.48 ± 15.3% vs 18.2 ± 13.6%; P < 0.001; the Society of Thoracic Surgery (STS) score: 6.4 ± 5% vs 5.5 ± 4.3%; P = 0.008). Valve predilatation was less frequently performed during VinV procedures (22.9% vs 69.1%; P < 0.001), and the hospital stay was shorter after VinV procedure (8.46 ± 4.2 days vs 9.83 ± 6 days; P = 0.005). VinV patients showed higher predischarge transvalvular mean gradients (14.14 ± 7.9 mmHg vs 8.42 ± 5.0 mmHg; P < 0.001), smaller mean valve surface area (1.54 ± 0.7 cm2 vs 1.83 ± 0.5 cm2; P < 0.001) and a lower risk of moderate/severe paravalvular leak (1.3% vs 5%). Post-procedural kidney injury (1.3% vs 4.8%; P = 0.06) and new pacemakers for conduction abnormalities (3.3% vs 18.5%; P < 0.001) were higher after TAVI. All-cause mortality and cardiovascular mortality at 30 days were similar between the 2 groups (1.9% vs 3.8%; P = 0.242 and 1.9% vs 3.4%; P = 0.321), whereas after 1 year, all-cause mortality was lower for VinV patients (6.8% vs 13%; P = 0.035). The bioprosthetic valve size correlated inversely with postoperative gradients after VinV procedures. VinV aortic procedures showed favourable 30-day and 1-year clinical outcomes compared with TAVI procedures for the native aortic valve disease. Despite higher transvalvular mean gradients following VinV implants, this appears not to impact the early clinical outcomes.