Predictors for new-onset conduction block in patients with pure native aortic regurgitation after transcatheter aortic valve replacement with a new-generation self-expanding valve (VitaFlow Liberty^™): a retrospective cohort study

Transcatheter aortic valve replacement (TAVR) has been preferred as the main treatment method for severe aortic stenosis (AS) owing to its safety and effectiveness [1‐6]. With growing expertise and significant clinical effectiveness in treating severe AS using TAVR, along with continuous technical improvements and advancements in device development [7‐9], the method is now widely employed in patients with anatomical diversity, including those with bicuspid aortic valves and failing surgical bioprosthetic valves [4, 5, 7, 10]. In recent years, researchers have been exploring the feasibility of TAVR for off-label indications, such as pure native aortic regurgitation (PNAR) [11, 12].

The accumulation of TAVR experience, coupled with improvements in technical parameters and valve devices, has contributed to the safety and feasibility of using TAVR in the treatment of PNAR [13, 14]. Compared with the first-generation valve devices, new-generation valve devices have been demonstrated to have better anchoring in the absence of significant aortic valve calcification in patients with PNAR [13‐15]. Numerous studies have shown that new-generation self-expanding valves can improve the success rate of TAVR in patients with PNAR and reduce complications, including paravalvular leakage and more [13, 14, 16‐19]. However, new-generation valves do not reduce the need for permanent pacemaker implantation (PPI) in patients with PNAR after TAVR [13, 14, 16‐20].

PPI is commonly required after TAVR, with registry studies reporting the use of PPI in 7.5–27.3% in patients with PNAR after TAVR [13,16,19,21,22]. New conduction block after TAVR seriously affects the prognosis of patients with aortic valve disease. A meta-analysis has shown that PPI after TAVR markedly increases the rates of all-cause mortality and rehospitalization due to heart failure [23]. The presence of a new postoperative left bundle branch block (LBBB) is an independent predictor for New York Heart Association functional class III or IV after TAVR for PNAR [15]. Ananwattanasuk et al. revealed that high-load right ventricular pacing or the presence of permanent LBBB after TAVR significantly increased the risk of 1-year postoperative mortality and cardiomyopathy [24].

Data from our hospital revealed that patients with PNAR were more prone to developing new conduction blocks requiring PPI compared to patients with AS [25]. Evidently, there is an urgent need for further research to investigate the causes of these complications and to optimize the management and treatment of patients requiring PPI after TAVR. Therefore, our study aimed to determine the clinical predictors of both PPI and new-onset conduction block after TAVR in patients with PNAR using a new-generation self-expanding valve (VitaFlow Liberty™, MicroPort, Shanghai, China).

A total of 88 patients with PNAR were treated with TAVR at our hospital from November 2021 to June 2023. Twenty patients were excluded from the study as follows: four patients with valve implantation failure due to an inability to anchor it intraoperatively, six had a preoperatively fitted pacemaker, one required transfer to surgery because the valve slipped into the LV while releasing it intraoperatively, eight had a non-VitaFlow Liberty valve, and one patient died in the hospital 48 h after the procedure due to a postoperative complication of thoracic aortic dissection. The remaining 68 patients met the inclusion criteria. Figure 1 presents the patient flow chart.

Out of 88 patients with PNAR, TAVR was unsuccessful in five cases, resulting in a success rate of 94.31%. Among the 83 patients who underwent a successful procedure, moderate or above-moderate AR was reported in four of them after the procedure, accounting for an incidence of 4.82%.

Twenty patients were newly implanted with a permanent pacemaker after TAVR, accounting for an incidence of 29.41%. Based on univariate analyses of baseline demographics, laboratory, electrocardiographic, echocardiographic, and CT measurement, and procedure data, and preoperative BNP levels, significant differences were noted in the presence or absence of preoperative complete right bundle branch block (cRBBB) or first-degree AVB and LVOT morphology between the pacemaker and non-pacemaker groups (Table 1). With inclusion of the three aforementioned significantly different indicators in the multivariate logistic regression analysis, presence of preoperative cRBBB or first-degree AVB [odds ratio (OR) = 31.44, 95% confidence interval (CI): 1.218–811.513, P = 0.038] and non-tubular LVOT (OR = 12.05, 95% CI: 1.124–129.199, P = 0.04) remained independent risk factors for pacemaker implantation (Table 2). Patients’ demographics and baseline data were also analyzed and compared between the two conduction disorder groups. Only LVOT type, aortic root valve angle, and valve implantation depth were significantly different between the two groups (P < 0.05; Table 3). After multiple logistic regression analysis that included the three aforementioned significantly different indicators, valve implantation depth (OR = 1.671, 95% CI: 1.189–2.348, P = 0.003) and angle of the aortic root (OR = 1.116, 95% CI: 1.012–1.230, P = 0.028) remained independent risk factors for cLBBB or high-grade AVB (Table 4).

Figure 6a presents the ROC curve for valve implantation depth as a predictor of cLBBB or high-grade AVB, using a cutoff value of 4.30 mm for valve implantation depth with a sensitivity of 61.9% and specificity of 90.9%. The area under the curve (AUC) of the ROC for valve implantation depth in predicting the new-onset of cLBBB or high-grade. AVB was 0.804 (95% CI: 0.666–0.942, P = 0.001). Further, Fig. 6b displays the ROC curve for angle of the aortic root as a predictor of cLBBB or high-grade AVB, using a cutoff value of 59° for the angle of the aortic root with a sensitivity of 71.4% and a specificity of 77.3%. The AUC of the ROC for angle of the aortic root in predicting the new-onset of cLBBB or high-grade AVB was 0.826, with a 95% CI of 0.570–0.882 and P = 0.011.

As the population ages, there is an increasing incidence of aortic valve disease. An epidemiological survey revealed that the prevalence of moderate-to-severe PNAR is approximately 1.2% in China, making it the most common type of valvular heart disease in the nation [30]. With the utilization of a new generation of self-expanding valves, combined with the growing experience of TAVR procedures and technological advancements, the success rate of TAVR in treating patients with PNAR has significantly increased. Moreover, many complications, such as PVR and moderate-to-severe AR, have been significantly reduced [13]. However, the occurrence of new PPI or new-onset conduction disorders, such as LBBB, has not been reduced in patients with PNAR after TAVR [13, 14, 16‐19], which seriously affects the prognosis of patients [14, 23, 24].

TAVR-related conduction disturbances, mainly new-onset LBBB and advanced AVB requiring PPI, remain the most common complication of this procedure [31]. Therefore, this study aimed to identify predictors of new-onset conduction block post-TAVR using new-generation self-expanding valves (VitaFlow Liberty™) in patients with PNAR. Our study revealed a 94.31% success rate of TAVR in patients with PNAR, with a 4.82% incidence of above-moderate AR after the procedure. Among patients with PNAR undergoing TAVR using self-expanding valves, preoperative cRBBB or first-degree AVB and a non-tubular LVOT were identified as indicators associated with a higher likelihood of requiring PPI. Additionally, deeper valve implantation depth and a greater angle of aortic root may serve as independent risk factors for new-onset cLBBB or high-grade AVB post-TAVR.

Chen et al. used first-generation self-expanding valves manufactured in China to perform TAVR in patients with PNAR and reported that VitaFlow™, which is a self-expanding valve with straight cylindrical shape (Fig. 5a), improved the success rate of TAVR compared with another self-expanding valve without a straight cylindrical shape [26]. Similarly, our results suggest that the use of VitaFlow Liberty™ can further improve the success rate of TAVR in patients with PNAR. The high success rate in our study might be attributed to the special aortic root anatomy of patients with PNAR and the self-expanding valve we used with a straight cylindrical shape (Fig. 5a). Typically, these patients have a large valve annulus without annular or valve leaflet calcification, which often necessitates anchoring the valve device in conjunction with LVOT [32]. The lower section of the VitaFlow™ valve device is essentially a straight cylindrical shape (Fig. 5a), offering increased contact area and friction between the valve device and the LVOT, thereby increasing the radial support force and facilitating valve anchoring [26]. Moreover, the new-generation VitaFlow™ valve system used in this study has recovery and repositioning functions, which also helps improve the success rate of TAVR in PNAR.

We found that preoperative cRBBB or first-degree AVB is an independent risk factor for PPI after TAVR in patients with PNAR, which is consistent with the results of a previous study on PPI after TAVR for AS [33]. However, to the best of our knowledge, this study is the first to show that non-tubular LVOT is also an independent risk factor for PPI in patients with PNAR after TAVR. We found that the non-tubular morphology of LVOT is mainly characterized by a flared or funnel shape. With a flared LVOT, the valve stent is prone to penetrate too deeply into the LVOT tract, leading to a conduction block. In contrast, with a funnel-shaped LVOT, the compression ratio of the lower valve stent edge is too large after successful implantation, and the radial force squeezing the outflow tract is too strong, leading to dysfunction of the bundle of His or LBB. Further randomized controlled studies are required to verify these results.

In addition, treatment of PNAR with TAVR poses unique technical challenges. Unlike patients with AS, most patients with PNAR do not have calcified valves and have a significantly dilated annulus and left ventricle, making fixation of the transcatheter valve prosthesis in the plane of the valve extremely difficult [34]. Thus, anchoring the transcatheter aortic valve prosthesis in patients with AS mainly relies on the aortic annulus, calcified valve leaflets, and STJ (Fig. 7a). In contrast to AS, the anchoring of transcatheter aortic valve prosthesis in PNAR relies on LVOT, aortic annulus, thickened aortic valve leaflet (if present), and STJ (Fig. 7b) [26, 32]. This may explain why, in our study, the risk factors for new PPI after TAVR differ between patients with PNAR and those with AS.

In line with previous studies, our study found a higher rate of new-onset conduction block in patients with PNAR after TAVR [13, 26]. The presence of new-onset conduction block after the procedure has significant effects on patient prognosis [14, 24]. In our study, greater aortic root angle and deeper valve implantation depth were independent risk factors for cLBBB or high-grade AVB. The importance of valve implantation depth was further supported by our ROC analysis findings, which are consistent with previous results on new-onset conduction block after TAVR for AS [33]. We also found a significantly increased risk of cLBBB or high-grade AVB with increased aortic root angle. The ROC analysis also further supported the predictive ability of aortic root angle for new conduction block after TAVR in patients with PNAR. The Youden index indicated that the sensitivity and specificity of predicting new conduction block post-TAVR were 0.714 and 0.773, respectively, when the angle of the aortic root was > 59°.

When performing TAVR in patients with PNAR using self-expanding valves, owing to the longer stent frame, the lowest point of the NCC serves as the initial supporting point for unfolding the valve stent device, and the valve is subsequently unfolded outward [35]. Finally, the prosthesis comes in contact with the side of the left coronary cusp (LCC), completing the valve release. Therefore, deployment of self-expanding valves is usually asymmetric [36]. Moreover, in our study, a smaller aortic root angle was associated with improved coaxiality and better alignment between the valve stent device and the aortic root, i.e., with a smaller angle between the plane of the lower valve stent device edge and the annular plane. In contrast, a larger aortic root angle was associated with worse coaxiality and poor alignment between the valve stent device and the aortic root, i.e., with a larger angle between the plane of the lower valve stent device edge and the annular plane and corresponding to an increased depth of the valve stent at the lower edge of the LCC and right coronary cusp (RCC). The close proximity between the heart conduction system and the aortic valve complex can explain the genesis of perioperative conduction disturbances during TAVR [37‐39]. Conduction disorders primarily result from direct mechanical damage to the conduction system, associated with ischemia, hematoma, and varying degrees of edema, occurring during the deployment of TAVR [40]. The triangle of Koch, defined by the ostium of the coronary sinus, the insertion point of the tricuspid valve septal leaflet, and the tendon of Todaro, is often used as an anatomical marker for the position of the atrioventricular node (AVN) within the right atrium [39, 41‐43]. The AVN continues as the bundle of His, penetrating to the left through the central fibrous body and piercing the membranous septum (MS). The position of the bundle of His is typically adjacent to the infra-anterior border of the MS, covered by a roof of ventricular muscle and ascending obliquely from the infra-posterior to supra-anterior directions [37, 39]. The MS is in conjunction with the RCC and NCC of the aortic valve [37]. There is considerable individual variability in MS length and its anatomical relationship with the bundle of His and the AVN [37]. Typically, the bundle of His penetrates the distal borders of the MS, adjacent to the transition to the muscular part of the IVS, and then separates into a left and right bundle branch underneath the surface of the LVOT [31, 39]. Kawashima and Sato showed that the area at greatest risk of injury during TAVR is the side of the MS adjacent to the RCC [37]. In summary, we can conclude that the His and LBB bundles are typically located in the LVOT below the RCC or at the LVOT below the junction of NCC and RCC. This positioning makes them susceptible to being compressed, potentially leading to a new-onset conduction block in the setting of TAVR. Thus, a greater aortic root angle corresponds to a deeper position of the valve stent at the lower edge of the LCC and RCC, which is consistent with the effect of valve implantation depth on the conduction system after TAVR. Based on the valve implantation depth, as defined in our study, an increased depth corresponded to a deeper position of the valve stent at the lower edge of the LCC and RCC. However, even at a normal valve implantation depth, a larger aortic root angle often results in a deeper device depth below the LCC and RCC (Fig. 4c and d). This may explain why a larger aortic root angle was identified an independent risk factor for cLBBB or high-grade AVB after TAVR in patients with PNAR in this study, which requires further validation in subsequent studies. However, to the best of our knowledge, this study is the first to show that the angle of the aortic root is also an independent risk factor for new-onset heart block in patients with PNAR who have undergone TAVR with self-expanding valves. Furthermore, this finding provides valuable insights into the need for optimizing the intervention process and the equipment used to enhance the coaxial alignment between the valve stent and the aortic root. Thus, we can effectively reduce the occurrence of new-onset heart block post-TAVR in patients with PNAR.

In summary, our study found that preoperative cRBBB, first-degree AVB, and non-tubular LVOT were closely associated with the need for PPI after TAVR in patients with PNAR. Moreover, the valve implantation depth and aortic root angle were closely associated with new-onset LBBB or high-grade AVB in these patients. Therefore, before performing TAVR in patients with PNAR, the possibility of new PPI or postoperative new-onset conduction block should be evaluated by integrating ECG, LVOT morphology, and aortic root angle measurements. Thus, based on the specific patient conditions, more optimal transcatheter interventional therapy strategies, such as valve implantation depth and valve device selection, may be planned.

When using a self-expanding valve for TAVR in patients with PNAR, preoperative cRBBB or first-degree AVB on ECG and non-tubular LVOT measured by TTE were associated with the need for new PPI. Further, intraoperative valve implantation depth and aortic root angle were also associated with postoperative new-onset cLBBB or high-grade AVB. Therefore, in cases wherein the preoperative ECG shows cRBBB or first-degree AVB, the LVOT is non-tubular, and the aortic root angle is large (i.e., horizontal heart), we should comprehensively assess whether to perform TAVR or optimize the procedure based on the patient’s condition, such as by minimizing the valve implantation depth under the NCC prior to valve device release, to avoid the occurrence of postoperative new-onset conduction block.