Comparison of a new bioprosthetic mitral valve to other commercially available devices under controlled conditions in a porcine model

Abstract Background/Aim To evaluate three mitral bioprostheses (of comparable measured internal diameters) under controlled, stable, hemodynamic and surgical conditions by bench, echocardiographic, computerized tomography and autopsy comparisons pre‐ and postvalve implantation. Methods Fifteen similar‐sized Yorkshire pigs underwent preprocedural computerized tomography anatomic screening. Of these, 12 had consistent anatomic features and underwent implantation of a mitral bioprosthesis via thoracotomy on cardiopulmonary bypass (CPB). Four valves from each of three manufacturers were implanted in randomized fashion: 27‐mm Epic, 27‐mm Mosaic, and 25‐mm Mitris bioprostheses. After CPB, epicardial echocardiographic studies were performed to assess hemodynamic function and define any paravalvular leaks, followed by postoperative gated contrast computerized tomography. After euthanasia, animals underwent necropsy for anatomic evaluation. Results All 12 animals had successful valve implantation with no study deaths. Postoperative echocardiographic trans‐valve gradients varied among bioprosthesis manufacturers. The 25‐mm Mitris (5.1 ± 2.7)/(2.6 ± 1.3 torr) had the lowest peak/mean gradient and the 27‐mm Epic bioprosthesis had the highest (9.2 ± 3.7)/(4.6 ± 1.9 torr). Surgical valve opening area (SOA) varied with the 25‐mm Mitris having the largest SOA (2.4 ± 0.15 cm2) followed by the 27‐mm Mosaic (2.04 ± 0.23 cm2) and the 27‐mm Epic (1.8 ± 0.27 cm2) valve. Bench device orthogonal internal diameter measurements did not match manufacturer device size labeling: 25‐mm Mitris (23 × 23 mm), 27‐mm Mosaic (23 × 22 mm), 27‐mm Epic (21 × 21 mm). Conclusions Current advertisement/packaging of commercial surgical mitral valves is not uniform. This study demonstrates marked variations in hemodynamics, valve opening area and anatomic dimensions between similar sized mitral bioprostheses. These data suggest a critical need for standardization and close scientific evaluation of surgical mitral bioprostheses to ensure optimal clinical outcomes.


| INTRODUCTION
Surgical valve design has undergone many iterations since 1952, when Charles Hufnagel implanted the first surgical valve, to treat aortic insufficiency. 1 Valve design has evolved to include development of mechanical, bioprosthetic, and ultimately rapid-deployment aortic valves for minimally-invasive approaches. There have also been major advances in the reduced need for anticoagulation, improved hemodynamic performance, and management of patient-prosthesis mismatch. However, as designs have evolved to tackle these challenges, there has been a lack of direct-independent scientific com-

| Study design and endpoints
Between August 2020 and January 2021, 15

| Animal preparation and examination
Before procedural consideration, all animals underwent anatomical evaluation with multidetector contrast-enhanced electrocardiographic (ECG) gated CT scanning, using an on-site Siemens scanner (Siemens Dual Somatom, Siemens Medical, Forchheim, Germany). 1,8 Preprocedural screening looked specifically at anatomical characteristics that would be used by a physician in the clinical setting. These data focused on evaluation of subjects' mitral annulus size during maximal diastolic dimensions, left atrial size and trans-septal catheter crossing height at mid-end systole. Those with transseptal crossing heights (defined as potential mid-mid transseptal fossa puncture to mitral annulus distance) ≤15 mm or mitral annulus dimensions (by diameters, area, or perimeter) with greater than 6% variation from other study animals were excluded from enrollment. Twelve pigs, all of similar physical size, met the inclusion criteria.
We performed bench measurements of prosthetic mitral bio-

| Data collection and statistics
Periprocedure (anesthetized but unoperated) multidetector retrospectively gated contrast enhanced CT scans were performed on all animals. Upon CT scan completion and dataset acquisition, multiphase cardiac reconstructions of the images were performed at 1.5 mm intervals. Images then were transferred in DICOM (Digital Imaging and Communications in Medicine) format for further evaluation. Postimaging CT processing was performed using Vitrea (Vital Images, Minnetonka, Minnesota) and Mimics (Materialise, Leuven, Belgium) software. All study animals were euthanized and underwent on-site supervised necropsy with cardiac explantation for anatomical evaluation of each surgical bioprosthesis. Given the small sample size, descriptive data is presented with no further statistical analysis.
Continuous and categorical variables are defined as mean and standard deviation; discrete variables are presented as numbers and percentages. 9

| RESULTS
A total of 15 animals underwent meticulous preclinical CT screening.
All animals were screened to obtain accurate size assessments of vascular structures, atria, ventricles, myocardium and mitral annulus.
Three animals were excluded due to annular size variations >6%, or transeptal crossing height ≤15 mm as evaluated by CT.

| Study population characteristics
For the 12 pigs, baseline age, body weight, left atrial size as well as annular anatomy are depicted in Table 1 Table 2.
Among the three studied bioprostheses, in descending order, the 27mm Epic mitral bioprosthesis had the highest peak/mean mitral gradient immediately post-implant, followed by the 27-mm Mosaic; the 25-mm Mitris had the least mitral peak/mean gradient (Table 3).
Doppler velocity indices parameters of all three mitral prostheses were within normal prosthetic mitral valve function parameters (Table 3).

| Risk of left ventricular outflow tract obstruction measures
There was no clinically significant LVOT obstruction in any study cases. After each successful surgical implantation, a 3D multiplanar CT reconstruction was performed to analyze the depth of the anterolateral and antero-septal struts within the LVOT ( Figure S1). The anterolateral strut of the 27-mm Mosaic bioprosthesis had the greatest strut depth (mean 11.3 ± 0.94 mm).
This was followed by the 25-mm Mitris device (8.6 ± 0.56 mm) and the 27-mm Epic with the shortest protrusion (8.4 ± 0.73 mm) (Table 4). This sequence remained similar for the antero-septal and posterior struts as well.
The depth of ventricular strut length protrusion did not correlate consistently with postsurgical mitral bioprosthesis LVOT gradients.
As mentioned, the 27-mm Mosaic had the greatest LVOT strut protrusion with the highest peak/mean LVOT gradient (4.4 ± 1.3)/ (1.9 ± 0.5 torr). However, despite having the shorter stent frame compared with the 25-mm Mitris, the 27-mm Epic trended toward having a higher peak/mean LVOT gradient than its counterpart, which had minimal change in this gradient from baseline (Table 4) ( Figure 3).

| Bench measurements: bioprosthesis frame internal dimensions
Among the three manufacturers' valves, measured internal diameters of new non-implanted bioprostheses demonstrated significant differences in valve frame design (Table S1 and Figure S2

| Surgical bioprosthesis strut design in the left ventricular outflow tract
Among the three types of surgical mitral bioprostheses, there was variation in strut length, strut width, and aortic outflow tract diameter opening between struts depending on strut location (Table S2). At the position of the aortic outflow tract, the 27-mm Mosaic had the longest and widest strut (14-mm length, 12-mm width), followed by the 27-mm Epic (8-mm length, 11-mm width) (  Here, we show that there was a strong connection between (1) SOA and mitral peak/mean gradient, (2) surgical internal frame design and mitral peak/mean gradient, and (3) (Table 4). However, the 27-mm Epic had wider strut dimensions (Table S2), and smaller distance between the struts oriented toward the aortic outflow tract, and a higher peak LVOT gradient  at the level of the tissue annulus, but additionally at the level of the ventricular struts. 3 There are conflicting reports on rates of structural valve deterioration, reoperation, and methodologies on how to assess hemodynamic function between bovine pericardial and porcine stented prosthetic mitral replacement. [13][14][15][16][17] Given the heterogeneity of mitral bioprosthesis valve design, there is a need for scientific standardization and validation of mitral pros-

| Limitations
This is a head-to-head early feasibility preclinical study on three specific surgical mitral bioprostheses. Several limitations to the study include the small number of animals studied and inability to test all surgical prosthesis sizes to justify certain valve type outcomes. Additionally, this is an acute animal study without ability to evaluate for long-term mitral bioprosthesis device durability. In this acute animal study, presence or absence of LV remodeling could not be considered in the hemodynamic evaluation of each mitral bioprosthesis. Although all efforts were made to control anatomical and hemodynamic variations, this study serves as a steppingstone for future human clinical studies. The importance of this study demonstrates the need for a pivotal trial with larger number of patients and longer period of follow-up to thoroughly assess the potential impact of surgical mitral bioprosthesis design on bioprosthesis function. Given the small number of animals in this pilot study, all results should be interpreted as hypothesis generating. Larger studies will be necessary to evaluate for long-term clinical outcomes.

| CONCLUSIONS
Rigorous scientific evaluation of surgical mitral bioprostheses is necessary for patient safety. Based on these results, we would advise caution when evaluating manufacturers' advertising. Implications of this study demonstrate a critical need for standardization and scientific evaluation of surgical mitral bioprostheses to ensure optimal outcomes for clinical human implantation.