Do Guideline-Based Indications Result in an Outcome Penalty for Patients With Severe Aortic Regurgitation?

ABSTRACT

OBJECTIVES: The present study examines whether improvements have reduced the negative impact of guideline triggers on postoperative outcomes. BACKGROUND: European and American guidelines for the management of severe aortic regurgitation (AR) define the triggers for AR surgery. These triggers are based on the results of studies performed in the 1990s analyzing outcomes of patients who underwent AR surgery in the 1980s. Although these triggers are used to indicate surgery, they have all been associated with poorer postoperative outcomes. In the meantime, innovations in operative techniques, including aortic valve repair, have allowed reducing the risk of surgery. METHODS: A total of 356 consecutive patients undergoing surgical correction of severe AR were included in this study. Among them, 204 were operated on for a Class I, 17 for a Class IIa, 49 for a Class IIb, and 86 without any guideline triggers. Cox proportional hazards regression models and Kaplan-Meier survival curves were used to compare postoperative outcomes in the different groups. Inverse probability weighing was used to adjust for mismatched baseline characteristics. RESULTS: Adjusted 10-year survival was better among patients without operative triggers (89 ± 4%) or with Class II triggers (85 ± 6%) than in patients with Class I triggers (71 ± 4%, p = 0.010). Similar results were obtained for cardiovascular survival and hospitalizations for heart failure. Spline function analyses indicated that mortality started to increase for left ventricular (LV) ejection fraction <55% and LV end-systolic dimensions >20 to 22 mm/m2. LV end-diastolic dimensions did not influence outcomes. CONCLUSIONS: Guideline-based Class I triggers for AR surgery carry major risks for long-term outcomes. This suggests that patients with severe AR should be operated on before the onset of these triggers; that is, at an asymptomatic stage, before LV ejection fraction falls below 55% or LV end-systolic dimensions exceeds 20 to 22 mm/m2.