Safety of percutaneous aortic valve insertion. A systematic review

Abstract

Background: The technique of percutaneous aortic valve implantation (PAVI) for the treatment of severe aortic stenosis (AS) has been introduced in 2002. Since then, many thousands such devices have worldwide been implanted in patients at high risk for conventional surgery. The procedure related mortality associated with PAVI as reported in published case series is substantial, although the intervention has never been formally compared with standard surgery. The objective of this study was to assess the safety of PAVI, and to compare it with published data reporting the risk associated with conventional aortic valve replacement in high-risk subjects.

Methods: Studies published in peer reviewed journals and presented at international meetings were searched in major medical databases. Further data were obtained from dedicated websites and through contacts with manufacturers. The following data were extracted: patient characteristics, success rate of valve insertion, operative risk status, early and late all-cause mortality.

Results: The first PAVI has been performed in 2002. Because of procedural complexity, the original transvenous approach from 2004 on has been replaced by the transarterial and transapical routes. Data originating from nearly 2700 non-transvenous PAVIs were identified. In order to reduce the impact of technical refinements and the procedural learning curve, procedure related safety data from series starting recruitment in April 2007 or later (n = 1975) were focused on. One-month mortality rates range from 6.4 to 7.4% in transfemoral (TF) and 11.6 to 18.6% in transapical (TA) series. Observational data from surgical series in patients with a comparable predicted operative risk, indicate mortality rates that are similar to those in TF PAVI but substantially lower than in TA PAVI. From all identified PAVI series, 6-month mortality rates, reflecting both procedural risk and mortality related to underlying co-morbidities, range from 10.0-25.0% in TF and 26.1-42.8% in TA series. It is not known what the survival of these patients would have been, had they been treated medically or by conventional surgery.

Conclusion: Safety issues and short-term survival represent a major drawback for the implementation of PAVI, especially for the TA approach. Results from an ongoing randomised controlled trial (RCT) should be awaited before further using this technique in routine clinical practice. In the meantime, both for safety concerns and for ethical reasons, patients should only be subjected to PAVI within the boundaries of such an RCT.