Computed tomography–based indexed aortic annulus size to predict prosthesis-patient mismatch transcatheter aortic valve replacement versus surgical aortic valve replacement in the SURTAVI trial

BACKGROUND: Hemodynamic performance of prostheses after transcatheter aortic valve replacement (TAVR) is generally better than after surgical aortic valve replacement (SAVR), especially in patients with a small native annulus size. However, it remains unclear whether differences are consistent for patients with a different propensity for developing prosthesis-patient mismatch (PPM), considering annulus size and body size of patients. METHODS AND RESULTS: The SURTAVI trial (Surgical Replacement and Transcatheter Aortic Implantation) compared TAVR using a self-expandable valve with SAVR in intermediate-risk patients. Multidetector computed tomography–based aortic annulus size consisted of the perimeter-derived diameter, which was divided by body surface area to produce an indexed annulus size. Patients were categorized into a small (9–12 mm/m²), medium (>12–14 mm/m²), and large (>14–18 mm/m²) group according to indexed annulus size. We compared TAVR and SAVR for PPM, hemodynamics, and clinical, and functional outcomes through 1-year follow-up within the size groups. Patients who underwent TAVR received a larger prosthesis with increasing indexed annulus size (P<0.001), while there was no difference in prosthesis size in patients who underwent SAVR (P=0.74). Patients in all size groups had significantly larger indexed effective orifice area and lower mean gradients at discharge after TAVR versus SAVR. Rates of PPM were significantly lower with TAVR versus SAVR in all groups (P<0.001) and declined with larger indexed annulus sizes with both TAVR (P=0.04) and SAVR (P=0.03). Indexed annulus size was an independent predictor of PPM after TAVR and SAVR. Clinical outcomes were comparable between TAVR and SAVR across all groups, apart from a significantly higher rate of reintervention after TAVR versus SAVR in the large indexed annulus size group (2.5% versus 0%; P=0.01) but without significant interaction (P_int=0.81). CONCLUSIONS: Rates of PPM were significantly lower after TAVR than after SAVR across all groups of indexed annulus size, reflecting better hemodynamic performance of transcatheter versus surgical valves, irrespective of the propensity to develop PPM. More attention should be directed to prevention of PPM after SAVR. This information should be considered by the Heart Team to recommend a specific procedure or valve.