Transcatheter aortic valve implantation and its impact on mitral valve geometry and function.

BACKGROUND
The aim of this study was to evaluate the impact of transcatheter aortic valve implantation (TAVI) on mitral valve geometry and function.


METHODS
Eighty-four patients underwent TAVI. Forty-four (52%) patients received a balloon-expandable valve and 40 (48%) were implanted with a self-expandable valve. All patients underwent three-dimensional-volumetric transesophageal echocardiography of the mitral valve before and immediately after TAVI. A dedicated software was used for assisted semiautomatic measurement of mitral annular geometry.


RESULTS
During systole, the anterior to posterior (AP) diameter was significantly reduced after the procedure (3.4 ± 0.5 cm vs 3.2 ± 0.5 cm; P < .05). The mitral annular area (10.8 ± 2.8cm2 vs 9.9 ± 2.6cm2 ; P < .05) as well as the tenting area (1.6 ± 0.7 cm2 vs 1.2 ± 0.6 cm2 ; P < .001) measured at mid-systole were reduced after TAVI. Diastolic measures were similar. Patients treated with balloon-expandable valves showed a significantly larger reduction in the AP diameter compared to self-expandable valves (-0.25 cm vs -0.11 cm; P < .05). The reduction of the annular area was higher in the balloon-expandable group (-1.2 ± 1.59 vs -0.22 ± 1.41; P < .05). Grade of mitral regurgitation did improve or remained stable after TAVI.


CONCLUSION
TAVI significantly impacts the mitral valve and mitral annular geometry and morphology. The choice of the prosthesis (balloon- vs self-expandable) may be relevant for those changes.


| INTRODUCTION
A relevant number of patients with severe aortic stenosis suffers from functional mitral regurgitation (MR). 1,2 Improvement in MR after aortic valve interventions has been described. 3,4 A common explanation for this phenomenon is the reduced end-systolic load and pressure immediately after aortic valve replacement in patients with aortic stenosis and the long-term positive effects of reverse remodeling. 5 Due to the proximity of the aortic and mitral valve annulus, mechanical interactions affecting the aortomitral curtain may also be relevant modifiers of this effect.
Changes in mitral annular geometry are known to affect the coaptation area and thereby function of the valve. 6 Changes in the degree of MR may also depend on the changes in mitral annular geometry after transcatheter aortic valve implantation (TAVI). 5    German Heart Center Berlin and were screened for this study. All patients were prospectively enrolled in the TAVI database of the German Heart Center Berlin. For the study, the database was screened for patients who had a three-dimensional (3D)-TOE before and after the intervention. Data extraction and image analysis were performed in a retrospective manner. All patients who had a TAVI procedure at the German Heart Center Berlin and had preinterventional and postinterventional 3D-TOE of the mitral valve with sufficient imaging quality and quantity were eligible for the study. Exclusion criteria were: previous surgery on the aortic valve and interventions on the mitral valve, such as ring annuloplasty and mitral valve replacement. Fourteen patients were excluded due to previous procedures affecting the aortic valve or the MV apparatus. In addition, 240 patients were not eligible due to lack of complete preoperative or postoperative 3D-TOEdataorlackofsufficient image quality. Additionally, 66 patients who underwent analgosedation with local anesthesia were excluded due to consecutive missing TOE examination. Hence, 84 patients were eligible for this study (Figure 1).

| TAVI procedure and echocardiographic assessment
TOE was routinely performed in cases with TAVI under general anesthesia. Besides the assessment of the aortic valve, evaluation of the mitral valve was performed during the standard TOE protocol. Imaging of the mitral valve was done preoperatively and postoperatively. The images were prospectively stored followed by image analysis at a later stage.

| Statistical analysis
Raw data from the ImageArena was exported and processed to allow comparability of the measurements. To deal with varying frame | 2187 numbers per cycle, the data were transformed as briefly described: The raw echocardiographic data consists of two variables per mea- indicating that mitral valve geometry undergoes substantial changes after TAVI procedures.  Figure 5).

| DISCUSSION
In this study, we could show that TAVI is associated with substantial changes in mitral annular geometry as assessed by 4D-echocardiography and measured with a semiautomated software tool. In contrast to the previous studies, which only examined systolic changes, 7-10 the observed effects were observed throughout the entire cardiac cycle; however, most of them did not differ significantly when compared to systole and diastole.
We found different changes at the mitral valve geometry, depending on the type of prosthesis implanted (balloon-vs selfexpandable). Changes were more relevant in the patients who had implantation of a balloon-expandable valve.
Previous studies investigating the effects on surgical SAVR also found significant changes in mitral valve geometry and function after surgical implantation of aortic valve prostheses. Mahmood et al 7,8 report in a single-center cohort study with 35 patients a significant reduction of mitral annular area, circumference, AP diameter, and  Note: Values are given as means and standard deviation. Abbreviations: ALPM, antrolateral-posteromedial; AP, anteriorposterior.

FIGURE 4
Changes of mitral valve geometry before and after transcatheter aortic valve implantation for all patients visualized over the whole heart cycle. A, Aortomitral angle (degree). B, AP diameter (cm); C, annulus area (cm 2 ). D, annular height (cm); E, sphericity index (cm/cm) Mahmood et al 7 on SAVR, also in our study, a less-steep aortomitral angle was detected after TAVI.
A study comparing the effects of TAVI and AVR on mitral annulus geometry found significant effects for AVR, but noted no changes for TAVI and, therefore, suggested that TAVI was a more physiological approach. 10 Investigations by Shibayama et al 9 found that TAVI did not affect the mitral valve annular 3D parameters, but did improve mitral leaflet tethering. In contrast to both the reports, our study showed a significant impact of TAVI on mitral annular geometry.
Many of the observed parameters such as AP diameter, annular area, and tenting parameters are known to be increased in patients with MR. 15,16 Changes in those annular dimensions are likely to have Abbreviations: ALPM, antrolateral-posteromedial; AP, anteriorposterior. While the ALPM diameter did not change significantly, the reduction of the AP diameter could be a surrogate for a mechanical component as an explanation for the observed changes, conceivable directly through the stent framework of the prosthesis pressing against the anterior mitral annulus. This hypothesis is further supported by the fact that significant differences were found between the groups of self-expanding vs balloon-expanded valves with a higher effect evident from balloon-expanded prostheses. However, hemodynamic changes due to the eliminated aortic stenosis such as afterload reduction may certainly also play a role. 18,19 Hence, we found possible causative factors for the observed changes, which still need to be validated in a larger patient cohort.

| LIMITATIONS
Our study was a single-center cohort study in which the collected data were retrospectively analyzed. Mitral calcification might affect the annular geometry but was not assessed in our study due to the assumption that calcifications affect the quality of imaging. Another aspect that most likely has an influence on mitral valve function and dynamics is the use of general anesthesia in all cases, but this might be less relevant because as a bias it equally affects all study patients. However, the exact influence of general anesthesia on mitral annular geometry is un-

| CONCLUSIONS
This study demonstrates that mitral valve anatomy and geometry is significantly influenced by TAVI procedures. This is even more pronounced for balloon-expandable valves and, thus, may be relevant for planning procedures and prosthesis selection. Grade of MR improved or was unchanged, which may also be attributed to the relevant geometric changes of the annulus in addition to the established mechanism of afterload reduction. To further assess the clinical relevance of these findings, further investigation in larger cohorts including a long-term follow-up is warranted. The use of magnetic resonance imaging for assessing mitral annular geometry could also provide additional information.