169-03: Echocardiographic Vs Non-invasive Hemodyanamic Optimization of AV and VV delay for Cardiac Resynchronization Therapy: The Prospective, multi-centre, randomized, cross-over, non-inferiority BRAVO study

Background: Landmark studies of biventricular pacing performed atrioventricular (AV) delay optimization, but the process requires considerable time of expert staff. BRAVO is a multi-centre, randomized, cross-over, non-inferiority trial comparing echocardiographic optimization of AV and interventricular (VV) delay with an alternative method using non-invasive blood pressure that can be automated to consume less staff resources. Methods: Patients with a previously implanted cardiac resynchronization devices were recruited and allocated to six months in each arm in random order. In the echocardiographic arm, AV delay was optimized by the iterative method, and VV delay by maximizing LVOT VTI. In the haemodynamic arm AV and VV delay were optimized using non-invasive blood pressure measured using finger photoplethysmography. At the end of each 6-month arm, the primary outcome was measured, objective exercise capacity quantified as peak oxygen uptake (VO2) during cardiopulmonary exercise testing. Secondary outcome measures were echo measurement of left ventricular remodelling, quality of life scores and NT-pro BNP. Summary of results: 401 patients were enrolled, median age 69 years, 78% male, NYHA class II 84%, III 16%. There was no significant difference in VO2 after 6 months with Echo optimized settings compared with hemodynamic optimized settings (mean difference 23 ml/min, P = 0.7). There was also no significant difference in LV volumes, symptoms or hormonal secondary endpoints: mean change in LV systolic dimension 1mm P= 0.18, LV diastolic dimension 0 mm (P = 0.93), Minnesota score -2 (P = 0.06), SF36 0.40 (P = 0.6), and NT-proBNP -10 pg/mL (P = 0.9). Conclusion: Optimization of AV and VV delay of CRT devices using non-invasive blood pressure is non-inferior to echocardiographic optimization. Therefore non-invasive haemodynamic optimization is an acceptable alternative, which has the potential to be automated and therefore more easily implemented.