The United Kingdom Transcatheter Aortic Valve Implantation (UK TAVI) Trial

DESIGN: The study comprises a prospective, multi-centre, open-label, parallel group, pragmatic, randomised controlled trial (RCT), comparing TAVI and AVR in patients with severe symptomatic aortic stenosis, who are at intermediate or high operative risk from AVR but considered suitable for either AVR or TAVI. SETTING: Specialist hospitals with active cardiac surgical and TAVI programmes, and at least 30 prior TAVI cases performed, to minimise learning curve effects. TARGET POPULATION: Patients with severe symptomatic aortic stenosis, referred for surgical intervention, aged 80 years or over, or aged 70 years or over with one or more factors associated with increased operative risk. The latter group is likely to comprise patients with STS risk scores of between 4% and 12% but these are not strict thresholds for inclusion and patients with higher or lower scores may be included if the MDT believes that there is clinical equipoise regarding the choice of intervention. HEALTH TECHNOLOGIES BEING ASSESSED: TAVI will be assessed as a generic intervention using any CE-marked device and any approach in current use, subject to satisfactory and sufficient global and local user experience, as adjudicated by the Trial Steering Committee. In the comparator group, any form of conventional surgical AVR will be accepted. MEASUREMENT OF COSTS AND OUTCOMES: The RCT will primarily assess non-inferiority of TAVI in respect of all-cause mortality at one year and then annually out to 5 years. Mortality is the primary consideration in considering operability and operative risk. The RCT will also assess the immediate impact of TAVI on quality of life. Long-term follow-up, out to 5 years, will include an assessment of the durabilty of any quality of life gains and the need for re-intervention. Cost-utility will be assessed using an NHS perspective over the duration of the trial and extrapolated over a lifetime using modelling techniques. Incremental cost-effectiveness will be calculated. SAMPLE SIZE: Assuming that there is truly no difference in one-year mortality between surgical AVR and TAVI, and that it is 15% for surgical AVR, allowing for 2% dropout in both arms, then 808 patients will be required to ensure 90% power that the upper limit of a one-sided 95% confidence interval will exclude a difference in favour of surgical AVR of more than 7.5%. The sample size will be reviewed by the Data Monitoring Committee in the light of the observed event rate in the surgical arm, and any treatment difference, and amended if necessary. PROJECT TIMETABLES: Planned study duration is 9 years, comprising 6 months set-up, 3 years recruitment, 5 years minimum follow-up and 6 months close-out. There are currently 20 experienced centres that have indicated their intention to participate. If 6 high-volume centres enrol 2 patients per month and 14 lower volume centres enrol 1 patient per month, full recruitment will be achievable within 3 years. Additional centres will further underwrite these conservative assumptions.