Prognostic influence of biological valve size on long-term outcomes of primary surgical aortic valve replacement for aortic stenosis.

Abstract

We clarified associations between biological valve sizes selected during primary surgical aortic valve replacement (SAVR) and its long-term outcomes. In total, 754 consecutive patients with aortic stenosis who underwent primary biological SAVR were classified into 3 groups based on valve size: 19-mm (n = 246), 21-mm (n = 262), and ≥23-mm (n = 246). Severe prosthesis-patient mismatch was defined as measured indexed effective orifice area ≤0.65 cm²/m² if body mass index is < 30 kg/m² and ≤ 0.55 cm²/m² if body mass index is ≥ 30 kg/m². The mean observation period was 6.6 ± 4.0 years (4996 patient-years). Patients with 19-mm valves were older (19-mm: 74 ± 7.8 years vs 21-mm: 72 ± 8.0 years vs ≥ 23-mm: 69 ± 10 years), predominantly women (91% vs 55% vs 13%), and had smaller body surface area (1.39 ± 0.12 m² vs 1.53 ± 0.15 m² vs 1.67 ± 0.16 m²). The incidence of post-SAVR severe prosthesis-patient mismatch was 15%, 5.8%, and 5.0%, respectively (P < .001 for all). However, a mixed-effects model demonstrated that left ventricular mass index regressed equivalently (interaction effect = 0.189). During follow-up, 142 (19%) mortalities and 41 (5.4%) aortic valve reinterventions (31 redo-SAVR and 10 valve-in-valve) were observed, without difference in 10-year all-cause mortality rate (21% vs 33% vs 27%; P = .438). Fine-Gray regression model identified age at surgery (adjusted hazard ratio, 0.4 per 10-years; P < .001) and use of a 19-mm valve (adjusted hazard ratio, 4.0; P < .001) as independent associates with aortic valve reinterventions. Despite a higher incidence of severe prosthesis-patient mismatch and aortic reintervention after primary SAVR, use of a 19-mm biological valve can be justified for elderly patients with small body surface area, as evidenced by equivalent long-term survival and left ventricular mass regression to those with larger valves.