Does treatment assignment influence the prognosis of patients with symptomatic severe aortic stenosis?

Objective: Aortic valve replacement (AVR) is the standard therapy in patients with symptomatic aortic stenosis (AS). In high surgical risk patients, alternative therapeutic options to medical treatment (MT) such as trans-catheter aortic valve implantation (TAVI) or balloon aortic valvuloplasty (BAV) have been proposed. In this study we evaluated whether treatment assignment influences per se the prognosis of these subjects. Patients and methods: Criteria for treatment assignment were based on patient's clinical conditions, Logistic EuroSCORE and other co-morbidities ignored by EuroSCORE. Due to baseline clinical differences between patients with diverse treatment assignment, we used propensity score matching to achieve balance. Results: 368 patients were studied: 141 underwent AVR, 127 TAVI, 49 BAV and 51 MT. 84 events (deaths for all causes) occurred during 14 months of follow-up: 11 AVR (8%), 26 TAVI (20%), 18 MT (35%), 29 BAV group (59%). Traditional Cox analysis identified treatment assignment as independent predictor of events (HR 1.82 [CI 1.10-3.25]) together with lower left ventricular ejection fraction, impaired renal function and history of heart failure. Matched Cox analysis by propensity score confirmed treatment assignment as an independent prognosticator of events (HR 1.90 [CI 1.27-2.85]), and showed similar rate events in TAVI and AVR patients, while it was significantly increased in BAV and MT patients. Conclusions: Treatment assignment may influence outcome of symptomatic patients with AS.