Rate, timing, correlates, and outcomes of hemodynamic valve deterioration after bioprosthetic surgical aortic valve replacement
Salaun E., Mahjoub H., Girerd N., Dagenais F., Voisine P., Mohammadi S., Yanagawa B., Kalavrouziotis D., Juni P., Verma S., Puri R., Coté N., Rodés-Cabau J., Mathieu P., Clavel MA., Pibarot P.
BACKGROUND: The incidence of structural valve deterioration after bioprosthesis (BP) aortic valve replacement (AVR) established on the basis of reoperation may substantially underestimate the true incidence. The objective is to determine the rate, timing, correlates, and association between hemodynamic valve deterioration (HVD) and outcomes assessed by Doppler echocardiography after surgical BP AVR. METHODS: A total of 1387 patients (62.2% male, 70.5±7.8 years of age) who underwent BP AVR were included in this retrospective study. Baseline echocardiography was performed at a median time of 4.1 (1.3-6.5) months after AVR. All patients had an echocardiographic follow-up =2 years after AVR (926 at least 5 years and 385 at least 10 years). HVD was defned by Doppler assessment as a =10 mmHg increase in mean gradient or worsening of transprosthetic regurgitation =1/3 class. HVD was classifed according to the timing after AVR: "very early," during the frst 2-years; "early," between 2 and 5 years; "midterm," between 5 and 10 years; and "long-term," >10 years. RESULTS: A total of 428 patients (30.9%) developed HVD. Among these patients, 52 (12.0%) were classifed as "very early," 129 (30.1%) as "early," 158 (36.9%) as "midterm," and 89 (20.8%) as "long-term" HVD. Factors independently associated with HVD occurring within the frst 5 years after AVR were diabetes mellitus (P=0.01), active smoking (P=0.01), renal insuffciency (P=0.01), baseline postoperative mean gradient =15 mmHg (P=0.04) or transprosthetic regurgitation =mild (P=0.04), and type of BP (stented versus stentless, P=0.003). Factors associated with HVD occurring after the ffth year after AVR were female sex (P=0.03), warfarin use (P=0.007), and BP type (P<0.001). HVD was independently associated with mortality (hazard ratio, 2.18; 95% CI, 1.86-2.57; P<0.001). CONCLUSIONS: HVD as identifed by Doppler echocardiography occurred in one third of patients and was associated with a 2.2-fold higher adjusted mortality. Diabetes mellitus and renal insuffciency were associated with early HVD, whereas female sex, warfarin use, and stented BPs (versus stentless) were associated with late HVD.