Prosthesis-patient mismatch after surgical aortic valve replacement with three different bioprostheses

A. D'Onofrio¹, G. Cibin¹, V. Lombardi¹, E. Bergonzoni¹, I. Cao¹, G. Lorenzoni¹, A. Francavilla¹, G. Gerosa^2
1University of Padova, Padova, Veneto ²University of Padova Medical School, Padova, Veneto

Purpose: Aim of this retrospective single-center study was to compare incidence and clinical impact of measured prosthesis-patient mismatch (PPMm) versus predicted PPM (PPMp) after surgical aortic valve replacement (SAVR) with three different bioprostheses. Furthermore we aimed at evaluating reliability of PPMp in identifying patients with postoperative actual PPM.

Methods: We analyzed data of all consecutive patients who underwent SAVR with Magna Ease, Inspiris Resilia and Intuity bioprostheses (Edwards Lifesciences, Irvine, CA) at our institution from 2016 until 2022. PPM was defined if EOAi ≤0.85 cm2/m2. PPMm was determined by institutional echo lab-measured EOAi on discharge day echocardiogram. PPMp was assessed using reference values for each valve model and size indexed to BSA based on height, weight, prosthesis type and size. For the overall population and for the three valve types we also evaluated sensitivity, specificity, positive predicted value, negative predicted value and accuracy of PPMp. Furthermore, consistency between PPMm and PPMp were evaluated according to prosthesis type, size, stent internal diameter (ID) and true ID. The clinical impact of PPMm and PPMp was evaluated calculating odds ratio (OR) for 30-day mortality by means of univariable logistic regression.

Results: During the study period a total of 1323 consecutive patients underwent SAVR at our institution. Complete hemodynamic data were available for 872 patients who represent the population of our study. Magna Ease, Inspiris Resilia and Intuity were implanted in: 446 (51.1%), 85 (9.7%) and 341 (39.1%) patients, respectively. The incidence of PPMp was 15.5%, 16.5% and 24% in Magna Ease, Inspiris Resilia and Intuity, respectively (p < 0.001). The incidence of PPMm was 22%, 20% and 10.3% in Magna Ease, Inspiris Resiia and Intuity, respectively (p < 0.001) (Fig. 1). In 635 out of 872 cases (72.8%) PPMp was consistent with PPMm (Magna Ease: 321/446, 72%; Inspiris Resilia: 58/85, 68.2%; Intuity: 256/341, 75%). Overall, sensitivity, specificity, positive predicted value, negative predicted value and accuracy of PPMp were: 0.26, 0.83, 0.24, 0.84 and 0.73, respectively (Magna Ease: 0.21, 0.82, 0.3, 0.8 and 0.72; Inspiris Resilia: 0.11, 0.82, 0.14, 0.79 and 0.68; Intuity: 0.45, 0.78, 0.19, 0.93 and 0.75). Table 1 shows consistency and discordance between PPMp and PPMm according to prosthesis type, size, stent ID and true ID. PPMm was significantly associated with 30-day mortality (OR: 1.89; 95%CI: 1-13-3.07; p=0.012) while PPMp was not (OR: 1.01; 95%CI: 0.57-1.71; p=0.96).

Conclusion: The incidence of PPMm and of PPMp was different among the three valve types. In particular, Intuity shows a significantly lower PPMm. Consistency between PPMp and PPMm was suboptimal especially in terms of sensitivity. Discordance between PPMp ad PPMm was more evident in smaller valve sizes. PPMm (but not PPMp) is significantly associated with 30-day mortality. A thorough knowledge of each valve type’s hemodynamics and of the reliability of predicting PPM especially in patients receiving small valves is crucial in order to select the most appropriate valve substitute and surgical technique (e.g. annular enlargement).

Identify the source of the funding for this research project: None