Optimal stroke volume index cut off for survival of severe aortic stenosis patients undergoing TAVI - where should we draw the line?

Abstract

Abstract Funding Acknowledgements Type of funding sources: None. Background Low-flow status in patients with severe aortic stenosis (SAS) is currently defined using stroke volume index (SVi), at a cut off below 35mL/m2. Low-flow has been associated with higher mortality in patients undergoing transcatheter aortic valve implantation (TAVI). However, different cut offs have been recently suggested to more accurately define a low-SVi associated with worse survival after valvular intervention. Purpose Determine the impact of SVi in mortality of SAS patients undergoing TAVI and define a prognostically relevant threshold in this context. Methods A database comprising all patients treated with TAVI between 2011 and 2019 in a single-centre was retrospectively analyzed, and cases with echocardiograms performed before intervention in our centre were included. Primary endpoint was defined as time to all-cause death or last follow-up over the five years after intervention. Surv_cutpoint from survival package in R was used to evaluate optimal low-SVi cutpoints associated with worse survival in SAS patients undergoing TAVI. Low-flow patients were compared with normal-flow counterparts using the determined low-SVi definition. The prognostic values of continuous and categorical SVi at a <35mL/m2 cut off and using the newly determined threshold were tested using Kaplan-Meier curves, log-rank test and Cox proportional hazard model adjusted for EuroSCORE II. Patients were further divided as having preserved or reduced ejection fraction (EF, <52%). p<0.05 was considered statistically significant. Results Of 657 TAVI performed, 488 (74.6%) patients were included in the present analysis, over a median follow-up of 56 months. A SVi<35ml/m2 was not associated with higher mortality after TAVI (p = 0.07). An optimal cutpoint of low-SVi was defined at <29mL/m2 (n=115, 24%), and these patients were on a more advanced NYHA class, were of higher estimated surgical risk, were more frequently hypertensive and anemic, and had a higher prevalence of atrial fibrillation. Echocardiographic data revealed lower EF, lower functional aortic valve area and lower transvalvular gradients in low-SVi patients. Regarding TAVI design, a balloon-expandable valve was more frequently used in this setting. SVi<29ml/m2 was associated with higher mortality after intervention [p = 0.003, hazard ratio (HR) 1.60 (1.18–2.17)], including after adjusting to EuroSCORE II, and in a reduced (but not preserved) EF context. When analyzed as a continuous variable, a higher SVi was associated with better survival after TAVI. Conclusions SVi is a prognostically relevant parameter in SAS patients undergoing TAVI. Contrary to a classically defined threshold of <35mL/m2, a SVi<29mL/m2 was associated with higher mortality after intervention in our population.