CT-based evaluation of patient-prosthesis mismatch after transcatheter aortic valve replacement (TAVI) and its influence on outcome

Background 

Patient-prosthesis mismatch (PPM) after transcatheter aortic valve implantation (TAVI) is a risk factor for heart failure and increased mortality. Transthoracic echocardiography (TTE) based assessment of PPM (TTE-PPM) tends to overestimate the severity compared to combining TTE-doppler-measurements with computed tomography angiography (CTA) left ventricular outflow tract diameter (CT-PPM). Furthermore, the presence of early asymptomatic valve thrombosis (hypoattenuated leaflet thickening, HALT) may lead to overestimation of PPM.

The aim of this study was to assess both methods for the incidence of PPM with an extent follow-up of more 5 years after exclusion of HALT. The study further aimed to evaluate anatomic predictors for PPM analyzing stent geometry in post-TAVI-CTA.

Methods and results

We evaluated the incidence of PPM in 444 patients(49.3% female, 82.0±5.3 years, 103 with self-expanding, 341 with balloon-expandable prosthesis). PPM was calculated using three different methods: 1. according the continuity equation from TTE measurements(TTE-PPM), 2. using the LVOT area from pre-TAVI-CTA(CT-PPM) and 3. in patients with self-expanding prosthesis using a neo-LVOT area (4mm below the new aortic annulus) from post-TAVI-CTA(neo-LVOT-CT-PPM). This was based on the assumption of a deeper LVOT-Doppler-velocity assessment in this prosthesis type.  PPM was defined as severe in patients with an indexed effective orifice area (EOA) of ≤0.65 cm²/m². Stent geometry (asymmetric expansion) was analyzed by measuring the prosthesis area at the LVOT end, the stent center and the aortic end, as well as the minimal and maximal diameter of the stent at each position respectively. Asymmetric expansion was then calculated as [(area stent entry + area stent exit)/(2 × area stent center)−1]×100. We further defined an eccentricity index as maximal diameter stent center/minimal diameter stent center. Patient’s clinical long-term outcome was evaluated over five years via standardized questionnaire.

A severe PPM was detected in 20 patients(4.5%) using TTE-PPM, in 2 patients(0.5%) using the CT-PPM and in 2 patients(1.9%) using the neo-LVOT-CT-PPM. Using multivariable linear regression analysis, body mass index (BMI) was identified as a predictor for PPM in each model (p<0.001). Using the TTE-PPM and CT-PPM, a smaller valve diameter was associated with the presence of PPM (p<0.001). Presence of a balloon-expandable prosthesis was associated with PPM using the CT-PPM only (p=0.033). Stent geometry did not influence the incidence of PPM (p>0.05 each method). Univariable cox regression analysis showed no influence of a severe PPM on overall survival, irrespectively of the method (p>0.05 each). Multivariable cox regression analysis showed a significant influence of a severe neo-LVOT-CT-PPM on overall survival after adjusting for BMI, valve diameter, valve type and stent geometry (p=0.037; Hazard Ratio 0.3).

Conclusion

We detected a very low incidence of PPM after exclusion of patients with HALT with even lower incidence of PPM by using the CT-based assessment. Only BMI was a predictor for PPM in each assessment model. PPM assessed by CTA was associated with a smaller valve diameter and presence of a balloon-expanding prosthesis. The incidence of PPM was not associated with stent geometry. Overall survival of these patients was significantly influenced by severe PPM assessed by using the neo-LVOT-CT-based method in patients with self-expanding prosthesis only.