CONCLUSIONS
A partner cohort study [8] showed that PVL (more than mild) was associated with an increase in the incidence of late mortality and severe PVL in the Pivotal Extreme Risk Trial. 9 A meta-analysis by Athappan et al.10 showed that in patients with moderate and severe PVL, the 1-year survival rate was reduced.
The presence of mild or more significant PVL after TAVR increases the incidence of mortality in the long-term.11,12 In the REPRISE III trial, the incidence of stroke events was increased in patients with mild PVL after TAVR. This indicated an association between PVL and stroke events. This raises the question of whether PVL may increase the risks of other postoperative complications.13 No practical method or uniformly used strategy for predicting PVL has been reported yet. It has been shown that one of the most significant predictors of paravalvular aortic regurgitation are valvular and annular calcium, which can predict PVL with a sensitivity and specificity of 86% and 80%, respectively, if the valvular calcium score was above 3,000.14 By applying such information, the sensitivity and specificity of the presence of cyclic calcium (i.e., asymmetric calcium) is a moderate predictor of Grade I or higher PVL.15 Reiff et al. reported that reproducing a high-fidelity model of the human aortic root was feasible in a clinical setting and did not require composite models and expensive printers. Furthermore, implanting a TAVR valve within a 3D printed model could aid predicting PVL. In the absence of adequate therapy for PVL, it is essential to develop a high-fidelity computational model to predict PVL.15 Revolutionary valve-implantation technology and improved preoperative planning with electrocardiography-gated multislice 3D-CT have overcome some complex problems16; however, PVL remains a problem in TAVR.17There is minimal PVL in conventional SAVR because the principle of SAVR is to remove as much calcification as possible in and around the annulus during SAVR. Many reports suggest that PVL does not affect mortality, cardiac accidents, deterioration of prosthetic valves, or artificial valve infection in the long-term.18,19 However, in addition to the above information, a meta-analysis of TAVR raised concerns that even mild perivalvular regurgitation can be detrimental to patient survival.20,21
RD-AVR has a low rate of PVL (severe aortic regurgitation(AR) 0.2%, moderate AR 1.2%, mild AR 6.1%), but these values are still higher than those reported after conventional SAVR interventions.22
The rate of significant postoperative PVL was 3.7% in the bicuspid aortic valve. These data are similar to those reported in previous studies. However, the risk of postoperative AR is 2.5 times higher than that of the total population of the SURD registry. We believe that it is essential to understand the complexity of the anatomy of bicuspid aortic valve (BAV) and restore the annular circumference for complete results.6, 22
To reduce PVL incidence, Andreas et al.23 reported the application ofan extra stitch in the noncoronary sinus in selected patients, and they did not implant this particular type of valve in patients with extensive calcifications of the root with rigid sinuses. Furthermore, SURD-IR data collection for more than 12 years showed that the incidence (17.8% to 2.7%)(Figure 1)and severity (severe AR 0.6% to 0, moderate AR 3.1% to 1.1%, and mild AR. 14.1% to 1.6%) of postoperative PVL decreased significantly. This may be due to the increased surgical experience in RD-AVR surgery.24
By observing preoperative data, Chirichilli et al. found that bicuspid aortic valves have a larger mean annular size in terms of area and perimeter compared to tricuspid aortic valves. This preprocedural anatomical finding might explain why preoperative CT scans showed an elliptic shape of TAVs (Ellipticity Index 1.3 ± 0.1), a circular shape of type 0 BAVs (1.1 ± 0.1) and an intermediate behavior of type 1 BAVs; this suggested a possible gradual spectrum of TAV circularity from type 1 BAV to type 0 BAV circularities.25 In our study, the 3D-CT reading was larger than the UCG reading, and the actual replacement prosthesis size was closer to 3D-CT tasks. Preoperative ellipse was not associated with perivalvular regurgitation, but perivalvular regurgitation was more common in patients with a large aortic annulus. This may be related to the large aortic annulus and the possible loss of active elasticity in the bicuspid valve.
We encountered four cases of PVL in the first half of this series, and one patient had hemolytic anemia, which required the replacement of the aortic valve. In our experience, regurgitation from the RCC-NCC commissure was the most common form, followed by regurgitation from the LCC-RCC commissure. It has been found that choosing a smaller prosthesis to avoid implantation of a new pacemaker or left bundle branch block increases the likelihood of having a gap in the commissure. If the height of the prostheticvalve is adjusted to the nadir, it will be in the infra-annular position at the commissure. Therefore, it was also found that a form with the commissure being high and spreading outwards had much regurgitation around the prosthetic valve. This morphology is especially common in bicuspid aortic valves. Ferrari et al. reported that the aortic annulus stabilization technique based on an already performed annuloplasty of the aortic annulus that uses a running 3–0 polypropylene suture was sufficient to prevent PVL.7However, our findings suggest that the elliptical shape of the annulus is not associated with PVL, and this is owing to a gap under the RCC-NCC commissure. Therefore, the additional stitch that we performed was easy and quick.
We changed the technique in the last 18 cases and inserted an everted mattress stitch at the commissure; the artificial valve was pulled up to the commissure and fixed. This extra stitch eliminated PVL complications, and there were no PVL incidences in the later 18 cases. We often choose a smaller-sized valve RD-AVR to avoid complete atrioventricular and left bundle branch blocks. In such cases, we are concerned about PVL, but we think this new stitch is a very effective way of eliminating PVL.
This study has some limitations. This was a retrospective case series with a small number of patients, meaning that there is also a risk of bias; there was also a potential for selection bias given that participants were all from a single university hospital. This did not allow us generalize the results of this study to different populations. Additionally, we reported only RD-AVR cases and did not show comparative data with conventional SAVR and TAVR. In future studies, we need to confirm the anatomical analysis of PVL for all aortic valve stenosis patients and long-term results and compare patients with and without PVL.
PVL is a rare complication of RD-AVR that may require reoperation and should be avoided. 3D-CT can measure PVL more accurately than UCG and is a useful test method for predicting PVL. We conclude that PVL is more likely to occur if there is a gap below the R-N commissure, especially in cases with large annulus sizes. In such cases, applying an additional stitch to the R-N commissure is extremely useful.
Acknowledgments: We would like to thank Editage (www.editage.com) for English language editing.