Abstract 19332: Results of Transfemoral Transcatheter Aortic Valve Implantation in Partner-Like Patients - A Comparison with Real-World Patients - The Bonn-Essen Experience

Abstract

Background: Transcatheter aortic valve implantation (TAVI) has rapidly evolved to the standard-of-care for inoperable patients with severe, symptomatic aortic valve stenosis and to an accepted treatment option for high-risk, but operable patients. However the randomized PARTNER trial excluded patients with conditions frequently encountered in our clinical practice (e.g. ejection fraction <20%, severe mitral regurgitation, recent coronary stenting, cerebrovascular accident within 6-months, bicuspid valve, severe aortic regurgitation, prior valve surgery, creatinine >3.0 mg/dl and/or hemodialysis). Methods: From 2006 to 2011, 485 high-risk patients (Age: 80 ± 7 years; AVA: 0.6 ± 0.2 cm²; log. EuroSCORE 22 ± 16%) underwent transfemoral TAVI (Edwards Sapien n=176, Medtronic CoreValve n=309). The patients were divided into a “PARTNER-like” group (n=272) and a “PARTNER-exclusion” group (n=213), on base of the published inclusion criteria of the PARTNER B cohort. Differences in perioperative, 30-day and 2-year survival were assessed by Kaplan-Meier survival analysis. Results: There was no statistical difference in patient-survival concerning site of treatment or the type of valve system. Patients in the “PARTNER-exclusion”- group had a worse two-year survival (Figure 1) than patients of the “PARTNER-like” group (83.8% vs. 77.5%, p=0.089). Also a non-significant trend towards a higher 30-day mortality (10.3% vs. 6.2%, p=0.102) was found in the exclusion group. The most frequent exclusion criteria were prior valve surgery (16%), followed by chronic renal failure (14%), severe mitral regurgitation (11%) and reduced ejection fraction (9%). Conclusion: In our two-center real world experience, TAVI patients who meet exclusion criteria for the PARTNER trial represent a cohort with higher long-term mortality. These findings emphasize the importance of optimal patient selection and might have implications on selection-strategies