Quantitative Analysis of Paravalvular Leak of Transcatheter Aortic Valves Using Cardiac MR: A Comparison of Surgical and Imaging Techniques

mation that is accurate and relevant to the Australian population can help with patient counseling, and may inform management decisions. We investigated the performance of commonly available risk-prediction scores, in predicting short-term mortality after surgery for endocarditis. Methods: Patients enrolled in the ANZSCTS database, undergoing endocarditis surgery from Jan 2001 to Dec 2011 were included. Pre-operative data were used to calculate All Procedures Logistic and Additive AusSCORE, EndoSCORE (a recently published 14-variable bedside score basedonSTSdata, targeted to endocarditis surgery), and Logistic and Additive EuroSCORE. Score performance in predicting 30-day mortality was assessed with ROC curves. Results: 1279 patients had surgery for endocarditis, with a 30-day mortality of 105 (8.2%). 1265 had complete data available for analysis. The performance of each score is indicated below. Score Area under curve 95% CI Additive STS EndoSCORE 0.782 0.739–0.824 Logistic AusSCORE 0.813 0.777–0.849 Additive AusSCORE 0.815 0.779–0.852 Logistic EuroSCORE 0.770 0.726–0.813 Additive EuroSCORE 0.779 0.737–0.821 Discussion: EndoSCORE performed well in predicting 30-day outcomes, and fitted Australian data with slightly better discrimination than in the STS cohort (AUC=0.758 for STS). The new All Procedures AusSCORE performed the best in predicting 30-day mortality after surgery for endocarditis in Australia. http://dx.doi.org/10.1016/j.hlc.2013.03.031 2012 Poster Presentation/Panel 6 Quantitative Analysis of Paravalvular Leak of Transcatheter Aortic Valves Using Cardiac MR: A Comparison of Surgical and Imaging Techniques G. Crouch1,2,∗, J. Bennetts 1,2, C. Bradbrook1,2, A. Sinhal 1,2, A. Penhall 1,2, R.A. Baker 1,2, J. Selvanayagam1,2 1Department of Cardiovascular Medicine, Flinders Medical Centre, Flinders University, Adelaide, Australia 2 Department of Cardiothoracic Surgery, FlindersMedical Centre, Flinders University, Adelaide, Australia Introduction: There is now significant registry and trial data demonstrating an increased incidence of paravalvular leakwith transcatheter aortic valve implantation (TAVI) than open aortic valve replacement (AVR). Despite improvements in both analysis and grading systems for echocardiography it largely remains qualitative in nature. We aim to provide quantitative analysis using CMR and compare it to concurrently collected echo parameters. Methods: Thirty-eight patients (20 male) with confirmed severe aortic stenosis undergoing either transcatheter valve replacement (20 patients) or high risk (euroSCORE>12) open AVR. CMR and transthoracic echocardiography (TTE) were carried out pre-operatively and within two weeks post-operatively. All CMR scans used a Siemens Aera 1.5T system and TTE a General Electric Vivid E9. Images obtained included standard cine functional imaging and aortic valve flow imaging. Results:Mean ages and log euroSCORE’s were similar. Post-operative CMR and TTEwere conducted at amedian of 6.0 days for TAVI and 7.3 days forOpen.Meanpreoperative left ventricular (LV) and right ventricular (RV) ejection fractions (EF) were similar in the two groups using CMR. Post-operative LVEFwas also similar in both groups. Postprocedure regurgitant fraction using CMR was higher in the transcatheter group (17.7 vs. 4.8% p< 0.01). Using published criteria for comparing CMR and qualitative echo, regurgitation was estimated at significantly lower values by echo than CMR (p< 0.01 see table below). A B S T R A C T S Heart, Lung and Circulation 469 2013;22:455–489 2012 ANZSCTS Annual Scientific Meeting None/Trivial Mild Moderate Mod-Sev Severe