A propensity matched analysis of outcomes and long term survival in stented versus stentless valves

Blake N. Shultz, Tomasz Timek, Alan T. Davis, John Heiser, Edward Murphy, Charles Willekes, Robert Hooker

To compare the perioperative and long term survival after aortic valve replacement using stentless versus stented valves in a large cohort of patients grouped using propensity score matching. From 1991 to 2012, 4,563 patients underwent aortic valve replacement with stentless and stented valves at our institution. Propensity score matching identified 444 pairs using 13 independent variables: incidence of operation, smoking status, renal failure, hypertension, diabetes, peripheral vascular disease, cerebrovascular disease, chronic lung disease, ejection fraction, gender, age, valve status, and use of coronary artery bypass graft. Data were collected from our Society of Thoracic Surgeons database and the Social Security Death Index. Groups were compared using univariate and Kaplan-Meier analysis. The two groups demonstrated no significant differences for the 13 matching variables and the majority of 30-day outcomes (p > 0.05). The stented valve group showed a higher incidence of postoperative bleeding (3.6% vs 1.1%, p = 0.015), but a lower incidence of stroke (0.9% vs. 2.9%, p = 0.028). One, five, and 10-year survival was 95.0, 80.7, and 52.8% for stented and 93.2, 80.5, and 51.3% for stentless valves. Overall survival did not differ significantly between the two groups (p = 0.641). Stentless and stented valves had identical 30-day outcomes except for a higher postoperative incidence of bleeding and a lower incidence of stroke in the stented group. There was no significant difference in long term survival between valve types. Both valves may be used for aortic valve replacement with low morbidity and excellent long term survival.