Background: Transcatheter aortic valve replacement (TAVR) has transformed the treatment of patients with symptomatic severe aortic stenosis (AS). Despite its widespread use there is only limited data on which hemodynamic variables can predict procedural outcomes and long-term mortality in these patients. Therefore, we evaluated the role of invasive left ventricular end-diastolic pressure (LVEDP) measurement as gold standard for diastolic dysfunction for predicting long-term mortality.

Methods: We performed a retrospective observational study on 3328 patients with severe aortic stenosis who had undergone TAVR at our center from July 2009 to June 2021 using a balloon-expandable (n=2197) or a self-expandable valve (n=1131). Based on invasive LVEDP measurement we retrospectively stratified the patients into two groups: patients with normal (< 15 mmHg) vs. elevated LVEDP (≥ 15 mmHg) prior to and after the procedure. The multivariate proportional hazard model was used to estimate hazard ratios (HR) and 95% confidence intervals (CI). Data collected included clinical, laboratory and echocardiographic variables. 

Results:

Elevated post-procedural LVEDP was identified in 2408 (72,3%) patients. The median LVEDP in the normal LVEDP group was 11.1 ± 2.8 and 21.5 ± 5.1 in the elevated group.The median clinical follow-up period was 36.6 months (IQR 6.4 – 40.3) months. All-cause in-hospital, 30-day, one-year, three-year and five-year mortality was 2%, 2.2%, 12.9%, 24.2%, and 36% respectively. 

The median overall survival was 61.9 months (95%CI 55.04-68.76) in the normal LVEDP group vs 54.1 months (95%CI 50.49- 57.65, p= 0.035) in the elevated LVEDP group. Five-year mortality rates were significantly higher among the elevated LVEDP group compared with the normal LVEDP group (5 years: 27.4% vs. 8.3%, p= 0.01; HR 1.22, 95% CI 1.05-1.41) (Fig. 1).

A multivariate model revealed the following independent predictors of five-year mortality after TAVI: postprocedural elevated LVEDP (HR 1.24, 95% CI 1.008-1.525, p=0.04), severe tricuspid regurgitation (HR 1.24, 95% CI 1.016- 1.516, p=0.03), pulmonary hypertension (PH) (HR 1.53, 95% CI 1.26-1.86, p=0.001). In the present study a significant paravalvular leak after TAVR was not associated with a higher mortality. (HR 1.45, 95% CI 0.95-2.19, p=0.082). 

Figure 1.  Adjusted all-cause survival curves for TAVR patients stratified by LVEDP

Conclusion: Elevated post-procedural LVEDP as read-out of diastolic function is a common finding in patients who underwent TAVR. Furthermore, it is also an independent predictor of all-cause mortality as well as PH and TR. This data confirms that diastolic dysfunction is an important predictor for mortality in TAVR and should be considered to guide procedure timing and management after TAVR.