ALSTER-TAVR 2024: Clinical Results at One Year Following Optimized Self-Expanding, Transcatheter Aortic Valve Replacement Employing the Cusp-Overlay Technique

© 2024 HMP Global. All Rights Reserved.
Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of the Journal of Invasive Cardiology or HMP Global, their employees, and affiliates. 

Objectives. Atrioventricular (AV) conduction disturbances are still common following transcatheter aortic valve replacement (TAVR). The study evaluates the feasibility and clinical effect of self-expanding (SE)-TAVR employing an optimized cusp-overlay technique (COT) at 1 year in a German all-comers population.

Methods. We analyzed 1-year clinical outcomes in patients who received a SE valve employing the optimized COT. Consecutive patients who underwent SE-TAVR (EvolutR, EvolutPRO; Medtronic) after introduction of the COT as the default implantation technique in 2020 were included (n = 101). Consecutive TAVR patients from the same operators using the conventional implantation technique (CIT) served as the control group (n = 116).

Results. The COT was successfully performed in more than 80% of the patients in the COT group. (81.2%, n = 82/101). At discharge, no difference regarding AV block of at least 2 (CIT 19.47% vs COT 21%; P = .86) and permanent pacemaker (PPM) implantation (CIT 17.5% vs COT 19%; P = .73) was observed between the cohorts. New left bundle branch block (LBBB) was significantly less frequent in the COT group (CIT 40.71% vs COT 26%; P = .029). Paravalvular leakage (PVL) greater than 1 degree was reduced in the COT cohort (CIT 8.62% vs COT 0.99%; P = .012). There was no significant difference in mortality (CIT 18.27% vs COT 13.83%; P = .44), stroke (CIT 9.62% vs COT 15.96%; P = .16) or cardiovascular rehospitalization after 1 year (CIT 25.96% vs 24.47%; P = .92) between the groups.

Conclusions. Implementation of COT-TAVR was feasible and safe, and it resulted in an improvement of TAVR outcomes regarding PVL greater than 1 degree and new onset LBBB. However, with respect to PPM, no difference was observed 1-year post-TAVR.

Recent randomized trials find transcatheter aortic valve replacement (TAVR) to be the standard therapy for aortic stenosis.¹ However, persistent rates of conduction system disturbances resulting in a need for permanent pacemaker (PPM) implantation are still affecting more than 20% of patients in recent trials.¹ A number of risk factors have been identified, including the use of self-expanding (SE) transcatheter heart valves (THV) and calcification of the left ventricular outflow tract (LVOT), as well as preexisting right bundle branch block (RBBB), left bundle branch block (LBBB), and left anterior hemiblock (LAHB).² In addition, implantation depth (ID) and a high Society of Thoracic Surgeons (STS) score were identified as independent risk factors for PPM. Patients who received a new PPM after TAVR are at risk for higher 1-year mortality, particularly those with an ejection fraction of less than 40%.³ We studied a cohort of consecutive TAVI patients (Figure 1).

The need for postprocedural PPM implantation results from complete atrioventricular (AV) block following the anchoring of the THV in the LVOT. The AV node is located in the triangle of Koch, which is defined by the tendon of Todaro, the orifice of the coronary sinus and the insertion point of the tricuspid valve. It proceeds to the His-bundle in the membranous septum and then diverts into the bundle branches.⁴ The left bundle branch proceeds to the triangle between the non-coronary (NCC) and right-coronary (RCC) cusps of the aortic valve. This close anatomical proximity is the reason for conduction disturbances in the context of TAVR. Conduction disturbances occur by direct mechanical compression from the TAVR prosthesis.

To minimize the interaction between the THV and the AV conduction system, particularly with SE valves (eg, EvolutR, EvolutPro; Medtronic), new implantation techniques were developed to allow minimal protrusion of the valve into the LVOT. This technique was termed “cusp overlay” (CO) since the primary anchoring of the THV is performed in an angulation where the body of the NCC serves as the orientation point. By overlapping the RCC and the left-coronary cusps (LCC) and isolating the NCC, a coplanar view is achieved. In addition, CO enables a good reference for the ID at the NCC, elongates the LVOT, and reduces the perceived distance of the THV frame between the NCC and the LCC (Figure 2A and B). The cusp-overlay technique (COT) has been attributed to less LVOT depth with SE-TAVR prosthesis.^5,6 Employing COT should lead to less interaction with the conduction system and is therefore expected to result in fewer PPM implantations.^7-9 Yet randomized, controlled trials comparing COT to conventional implantation techniques (CIT) are lacking; furthermore 1-year follow-up data are not yet available. We previously reported on our experience switching a TAVR program from mainly balloon-expandable (BE) valves to SE valves for cost effectiveness.¹⁰ Here, we studied the safety, feasibility, and clinical outcomes at 1 year following the 2020 implementation of the COT compared to the previously studied CIT.¹⁰

The ALSTER TAVR 2024 study is a retrospective, multicenter registry with the purpose of collecting clinical evidence on valve performance and procedural and clinical outcomes in a German all-comers population. ALSTER refers to the department (ASKLEPIOS ALTONA DEP. of CARDIOLOGY) as well as the Hamburg city lake (“ALSTER”), following the same naming convention as our previous studies (ALSTER-TAVI 2021, ALSTER-OCT-CTO 2016, ALSTER-Left Main 2016, ALSTER BP 2014, ALSTER-Stem Cell 2012). This analysis focuses on the effect of introducing the COT for the purpose of achieving the ideal LVOT depth to anchor the SE Evolut valve type. Data were compared to a similar patient cohort that was studied prior to the introduction of the COT in mid-2020 (Figure 1). Data were analyzed as an intention-to-treat approach. The ALSTER TAVR study adheres to the ASKLEPIOS Hamburg institutional guidelines and, as a retrospective analysis, consent was waived by the local body of our institution.

Study population. The COT group included all patients undergoing transfemoral TAVR between June 2020 and October 2021, during which time the COT became the standard of care. In the COT group, 81.2 % (n = 82/101) of TAVR were implanted using the COT; 18.8% (n = 19/101) were implanted using the CIT due to anatomical or procedural restrictions (eg, angulation, anchoring) that prevented the operator from using the COT. We use the CIT group as the control; this group was the previously analyzed patient cohort who underwent TAVR between January 2016 and August 2018 (ALSTER TAVI).¹⁰ Patients with indications other than severe aortic stenosis, such as bicuspid valves or valve-in-valve procedures, or patients who received BE valves, were excluded. The selection of valve type was at the discretion of the operator, with the SE-TAVR Evolut platform being the default strategy (> 80%) for cost effectiveness. Because we aimed to study the influence of COT on new conduction disturbances and PPM implantation, we excluded patients with preexisting PPM (Figure 1).

In total, 116 CIT-TAVR patients and 101 COT-TAVR patients were analyzed. Baseline and periprocedural data were retrospectively collected from patients' charts and discharge letters. Clinical follow-up information was obtained by phone from referring physicians or hospital admissions. Clinical endpoints and periprocedural complications were defined with the updated Valve Academic Research Consortium (VARC)-3 definitions.¹¹ Follow-up at 1 year, including echocardiographic assessments, was available from more than 90% of patients from each group.

Multislice computed tomography imaging. Multislice computed tomography (MSCT) was performed in all patients to show the native aortic valve with dimensions of annulus and LVOT, as well as to assess the feasibility for transfemoral TAVR access. In the COT group, calcification was quantified at various levels by employing the dedicated 3mensio Structural Heart aortic valve software (Pie Medical Imaging BV). The aortic annulus diameter was defined as a virtual plane with the basal attachment points of the 3 leaflets; LVOT diameter was defined as 5-mm inferior to the annulus (Figure 2A).¹² To discriminate calcification and contrast medium, 550 Hounsfield Units were set. The supra-aortic area was defined as between the annulus plane and the coronary ostia (Figure 2A).¹³ The annulus area was set at 2-mm inferior and 3-mm superior of the annulus plane, the LVOT area at 5-mm inferior of the annulus plane, and leaflets at 3-mm superior of the annulus plane to the edge of the cusps as described previously (Figure 2A).¹² In the angiographic view, the conventional 3-cusp and CO angulation was simulated to predict the angulation of the CO during the procedure (Figure 2B).

TAVR procedure. The diagnosis of severe aortic stenosis was made according to the current European Society of Cardiology (ESC) guidelines.¹⁴ The decision for a TAVR procedure vs an open-heart surgery was made by the Heart Team and was mostly limited to high-risk (Euro Score log > 6) or elderly (> 75 years) patients, reflecting the current German standard. Selection of valve type was at the discretion of the operator, with the SE EvolutR or EvolutPro being the default strategy for cost effectiveness. Valve size was selected based on the manufacturer instructions. During the time of the registry, both SE EvolutR (CIT n=110, COT n = 85) and EvolutPRO (CIT n = 6, COT n = 16) were utilized.

Transfemoral access was performed in all patients who were analyzed in this study. The decision for pre- or post-dilatation was at the discretion of the operator.

The inner diameter was measured retrospectively on x-ray documentation before and after valve release. The diameter was measured in an angulation that allows a symmetric 3-cusp view. In the COT group, the depth was additionally measured in CO view before valve release. Measurements were performed in the Dicom Viewer OsiriX TM Version 3.3.1 (Figure 3).

Study endpoints. The primary aim of the study was to understand the impact of COT-TAVR on conduction disturbances and the need for PPM implantation, as well as the clinical outcomes at 1 year in an elderly German all-comers collective. Over 85% of patients received PPM implantation due to complete heart block. As secondary outcomes, we evaluated the procedural endpoints according to the updated VARC-3 definitions and the heart failure symptoms as defined by New York Heart Association (NYHA).¹¹

Statistical analysis. For descriptive statistics, continuous variables are shown as median with interquartile range (IQR) due to non-normal distribution. Binary variables are presented as absolute numbers and percentages. To compare the 2 groups for continuous variables, the Mann-Whitney test or unpaired Student’s t-test was employed. Fisher’s exact test was used for binary variables due to limited sample size. A 2-sided P-value less than .05 was considered to be statistically significant. Survival curves were produced using the Kaplan-Meier method. Statistical analysis was performed using GraphPad Prism, version 10.1.1 (GraphPad Software).

Patient baseline characteristics. Baseline characteristics were well matched regarding age (CIT 83.4 years [86.3, 80.4] vs COT 83.3 [86.3, 80.2] P = .96), female gender (CIT 53.5% vs COT 54.5%; P = .89), and STS score (CIT 3.4 [2.6, 4.4] vs COT 2.9 [2.1, 4.8]; P = .16) or relevant comorbidities. Baseline conduction disturbances, especially preexisting RBBB (CIT 8.6% vs COT 9.9%; P = .82) and LAHB (CIT 22.41% vs COT 14.9%; P = .17) were similar. The EURO score log was significantly higher in the CIT group (CIT 15.7 [10.8, 22.8] vs COT 12.98 [8.5, 20.0]; P = .02) while NYHA functional class (CIT 3 [2, 3] vs COT3 [3, 3]; P < .001) and NT-proBNP (CIT 1195 ng/L [690, 2434] vs COT 1778 ng/L [811.3, 3886]; P = .004) were significantly higher in the COT group (Table 1).

Procedural data. In the COT group, 82 of 101 patients (81.2%) were implanted according to the COT standard, while 19 (18.8%) patients demonstrated an anatomic variation that precluded the operator from using the COT (ie, the necessary angulation would exceed 40° caudal or several attempts of LVOT anchoring in the CO angulation were unsuccessful). In these cases, the conventional 3-cusp view (CIT) was used instead. Procedure time (CIT 65 minutes [57, 79] vs COT 63 minutes [51, 77.5]; P = .2) fluoroscopy time (CIT 13.4 minutes [10.9, 17.2] vs COT 13.2 minutes [10.5, 17.1]; P = .65) and amount of contrast medium (CIT 153 mL (110, 185) vs COT 150 mL (120, 181.5); P = .74) between the groups did not differ. No valve embolization was observed.

Valve sizing was performed according to the manufacturer instructions based on anulus perimeter-derived size from MSCT. The EvolutR/PRO 29-mm valve size was the most common valve (CIT 50%, n = 58 vs COT 50.5%, n = 51) employed, followed by the 34-mm size (CIT 24.1%, n = 28 vs COT 24.8%, n = 25). Valves sized 26 mm were implanted in 20.7% (CIT n = 24) and 21.8% (COT n = 22) patients, respectively. The EvolutR/PRO 23 mm was used only in 5.2% (CIT n = 6) or 3% (COT n = 3) of all TAVR patients. Oversizing (CIT 20.6% [13.0, 26.1] vs COT 18.9% [13.3, 23.7]; P = .59) was very similar in both groups. Pre-interventional RBBB was observed in approximately 9% of the patients (CIT 8.6% vs COT 9.9%; P = .82) Pre-dilatation was performed significantly more often in the COT group (CIT 49.1% vs COT 70.3%, P = .0023), while post-dilatation was the same (CIT 37.1% vs 30.7%; P = .39). The ID of the THV under the NCC was significantly less in the COT group (CIT 6.43 mm [4.9, 7.86] vs COT 5.07 mm [3.5, 6.76]; P = .0002) before release. Yet, after release from the delivery system, the ID under the NCC (CIT 7.35 mm [4.46, 9.85] vs COT 6.33 mm [3.87, 9.14]; P = .24) and the averaged ID to the NCC and LCC (CIT 8.07 mm [6.39, 10.37] vs COT 7.36 mm [5.76, 10.25]; P = .19) was numerically but not statistically significantly less in the COT cohort (Table 2, Figure 3).

Conduction disturbances and pacemaker. There was no significant difference between the rate of complete heart block (CIT 19.47% vs COT 21%; P = .86) and the rate of PPM implantation in-hospital (CIT 17.5% vs COT 19%; P = .73), after 30 days (CIT 20.75% vs COT 20.83%; P > .99), after 6 months (CIT 23.81% vs COT 22.92%; P > .99) and after 1 year (CIT 26.92% vs COT 24.47%; P = .83). Upon stratifying the results for the different valve sizes, we observed a significantly lower rate of PPM implantations in the COT group if the patients received a 26-mm valve (CIT 29.17% vs COT 4.55%; P = .049; Figure 4). In addition, the rate of postinterventional LBBB (CIT 40.71% vs COT 26%; P = .029) was significantly lower in the COT group (Figure 4). The percentage of patients with a need for PPM after discharge from the hospital up to 1-year follow-up was numerically lower in the COT group (CIT 9.4%, n = 8 vs COT 5.5%, n = 4; P = .33). While more than 50% of patients in the COT cohort presented with LVOT calcification (Table 1), this was not predictive of PPM in this cohort. Surprisingly, we observed calcification at the annular level (COT-PPM 110 mm³ [46.5, 777] vs COT-no-PPM 83 mm³ [29, 186]; P = .017) to be predictive of PPM in this study (Table 2, Figure 5).

Clinical outcome. The rate of paravalvular leakage (PVL) greater than I° was significantly lower in the COT group (CIT 8.62% vs COT 0.99%; P = .012) (Table 3, Figure 4). There was no significant difference in periprocedural mortality (CIT 1.72% vs COT 0.99%; P = .64), in-hospital mortality (CIT 3.45% vs COT 1.98%; P = .51), non-disabling (CIT: 1.72% vs COT 5%; P = .26) or disabling stroke (CIT 1.72% vs COT 0.99%; P > .99), bleeding (CIT 21.55% vs COT 13.86%; P = .16), or vascular complications (CIT 10.34% vs COT 3.96%; P = .12) in-hospital. The duration of hospital stay was lower in the COT group (CIT 8 days [7, 10] vs COT6 [5, 8]; P < .0001). At the 1-year follow-up, 15 patients from the CIT group and 11 from the COT were found to have died (CIT 18.27% vs COT 13.83%; P = .44). There was no significant difference in disabling stroke (CIT 2.88% vs COT 3.19%, P = .86), TIA/non-disabling stroke (CIT 6.73% vs COT 12.77%; P = .23), or rehospitalization rate for a cardiovascular cause (CIT 25.96% vs COT 24.47%; P = .92) (Figure 6). Interestingly, COT patients reported a significantly lower NYHA class (CIT 2.24 [1.92, 2.53] vs COT 1.86 [1.63, 2.28]; P = .0002) at 1 year despite NYHA class and NT-proBNP levels were higher at baseline (Table 3, Figure 4).

In our study comparing 2 well-matched German all-comers cohorts of more than 100 TAVR patients with a mean age of greater than 80 years, we found that switching from CIT to COT SE-TAVR was feasible in more than 80% of patients. No safety concerns occurred aiming for less LVOT depth employing COT. There was no difference in mortality, stroke, or rehospitalization after 1 year between the 2 cohorts. The COT aims to optimize the implantation depth with the SE Evolut TAVR platform. Yet, COT did not affect the rate of PPM due to an AV block greater than 2 at 1 year in our cohort. However, PVL greater than 1 degree and the occurrence of new LBBB decreased in favor of COT. While baseline heart failure burden as determined by NT-proBNP levels and NYHA classification was higher in the COT cohort at baseline, these patients had less heart failure symptoms as classified by NYHA at 1 year. This observation was associated with low aortic valve gradients following SE-TAVR and less patients with new LBBB in the COT cohort at 1 year.

The 2 cohorts were well matched regarding age, STS score, and gender distribution. They reflect an elderly and high-risk collective. While the Euro score log shows that the high-risk, STS score may appear relatively low, this is explained by the more complex calculation and the need for more data, as well as the retrospective character of the study. Valve oversizing was similar and compares well to other recent studies comparing CO-TAVR to the 3-cusp TAVR.¹⁵ There was no significant difference between preexisting RBBB or LAHB, first-degree AV block, and perimeter-derived annulus diameter between the 2 cohorts; these findings were predictive of new PPM implantation rates. Like other series, there was no safety concern regarding CO-TAVR. The duration of hospitalization was lower in the COT group. In general, TAVR in Germany is associated with hospitalization for 6 to 16 days, in part due to extensive in-hospital mobilization and stabilization prior to discharge.^16,17 While newer studies have shown the safety of earlier discharge with prespecified risk criteria, this has not yet translated to daily practice.

This trial is one of the first data sets of the COT with a follow-up beyond 30 days: we observed no increase in valve embolization and PVL decreased significantly, probably due to the optimized ID, which maximizes the radial force of the SE THV employing the COT.¹⁸ Less than 15% of the COT patients received the EvolutPRO valve that incorporates a LVOT skirt; reduced PVL is therefore not explained by the skirt, yet is associated with COT. Like age-matched populations that have been studied in various other registries including intermediate and high-risk patients, 1-year mortality was high in both cohorts, at around 16%.^19,20 Stroke has been described at a rate of around 17% at 5 years;²⁰ here, we observed a rate of 9.6% (CIT) vs 13.9% (COT), respectively. There was no difference between the groups regarding rehospitalization at 1 year.

Procedure planning included CT assessment of the annulus and predefining the angulation to achieve CO implantation. Assessment of the TAVR LVOT ID found a significant difference only prior to release of the valve, while the final ID under the NCC as measured in a final 3-cusp view was only numerically different between the 2 cohorts. We believe this to be due to self-orientation of the valve following release in a 3-dimensional context leading to final anchoring. Absolute values in our series are higher than data from a large meta-analysis recently published, wherein the distance from the NCC hinge point to the ventricular end of the THV frame was measured with 4.5 mm vs 5.5 mm comparing CO-TAVR to CIT-TAVR.¹⁵ Yet only 3 of 5 trials measured the ID with huge variations regarding the final angulation employed for measurement. In addition, 2 of 5 trials included only propensity-matched COT patients, while our study analyzed 2 all-comer patient cohorts. ¹⁵

A recent meta-analysis of 1227 patients from 5 observational studies on the effect of COT-TAVR found a significant reduction for PPM implantation compared with a series where TAVR with SE valves was performed using CIT. Only 9.8% of COT-TAVR patients were in need of a PPM postprocedural, compared with 20.6% of CIT-TAVR patients.¹⁵ We observed a significantly lower rate of PPM implantations with the COT only in patients who received a 26-mm valve (CIT 29.17%, n = 7 vs COT 4.55%, n = 1; P = .049), while the overall PPM rate was unchanged: 26.9 % (CIT) vs 24.5% (COT) at 1 year. This finding is associated with a high rate of pre-existing conduction disturbances in our cohort; more precisely, 20% of patients demonstrated a first-degree AV block and a 19% LAHB prior to TAVR, compared with around 10% in other series.¹⁵ In addition, more than 50% of our patients demonstrated LVOT calcification predictive of PPM, compared with less than 10% in other series studying the COT.⁷ These findings are consistent with a recent publication that also compares the COT with the CIT regarding PPM rates at hospital discharge: this study included 260 patients in each cohort and, in both groups, around 17% of patients required PPM implantation.²¹ Interestingly, we found annular calcification to be predictive of conduction system abnormalities following TAVR, a risk factor that might explain the observation that, even with CO-TAVR, the PPM rate due to AV-block 3 is still high.^8,22

Patients receiving TAVR seek reduction of their heart failure symptoms. A recent study analyzing multiple domains of cardiac function, hemodynamics, and exercise capacity show that HFpEF status is an important factor in addition to aortic stenosis severity.^23,24 Despite higher baseline NT-proBNP and NYHA class in the COT cohort, heart failure symptoms improved statistically significantly with CIT (P < .001). This observation is associated with improvements regarding remaining PVL, less patients with new LBBB, and a physiologically low gradient over the SE-TAVR prostheses following implantation. We previously observed a significantly lower transvalvular gradient following TAVR employing the EvolutR or EvolutPro compared with the balloon-expandable SAPIEN 3 at 12 months, yet without significant clinical differences in more than 100 patients in each cohort.¹⁰ Further data are needed to confirm that SE-valve implantation by COT is associated with less heart failure symptoms.

Limitations. We analyzed 2 all-comers cohorts where the COT was feasible in more than 80% —but not all —patients, while other recently published studies focused on a highly selected patient collective with the COT applied in 100% of patients. While the team of TAVR operators and their preplanning routine remained unchanged between the cohorts, the retrospective design leads to less standardization regarding final angulations and other procedural details. Data was collected from the beginning of the implementation of the COT, so the learning curve with the new technique is included. The difference to other studies in the field describing a reduction of PPM with COT might be explained by our study including high-risk, elderly patients with multiple comorbidities, an already “high” implantation employing CIT before switching, and no significant difference of LVOT depth in the 3-cusp view after COT implementation.

In summary, SE-TAVR implantation has evolved during recent years with the COT becoming standard. COT aims to reduce PPM rates by minimizing LVOT ID. Albeit ID was only numerically reduced, this study finds COT to improve heart failure symptoms at 1 year associated with less PVL > I° and number of patients with new LBBB compared to a well-matched conventional SE-TAVR patient cohort. However, new PPM rates following SE-TAVR were not significantly reduced, likely related to a high rate of patients with pre-existing conduction disturbances plus both LVOT as well as annulus calcification in MSCT measurements of our COT cohort. Further modifications to valve design and patient screening are necessary in order to minimize PPM rates following SE-TAVR.

COT-TAVR with EvolutR/EvolutPro is feasible and safe, but did not significantly affect the final ID and PPM rate in this all-comers series of high-risk patients with an average age of over 80 years. Clinical results regarding heart failure symptoms improved with COT-TAVR compared to CIT-TAVR at 1 year. In addition, COT-TAVR is possibly linked with less PVL greater than 1 degree, low post-procedural aortic valve gradients, and less patients with new onset LBBB following TAVR.

From the ¹Department of Cardiology and Intensive Care, AK Altona, Hamburg, Germany; ²Interventional Cardiology, Cardiologicum, Hamburg, Germany; ³University Heart & Vascular Center Hamburg, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; ⁴German Center for Cardiovascular Research (DZHK), Partner Site Hamburg/Kiel/Lübeck, Hamburg, Germany; ⁵Department of Cardiology, Universitätsklinik Schleswig-Holstein, Campus Lübeck, Germany; ⁶University Heart Center Lübeck, Department of Rhythmology, University Hospital Schleswig- Holstein, Germany; ⁷Department of Cardiology, AK St. Georg, Hamburg, Germany; ⁸Department of Cardiothoracic Surgery, AK St. Georg, Hamburg, Germany.

Disclosures: The authors report no financial relationships or conflicts of interest regarding the content herein.

Address of correspondence: Martin W. Bergmann, MD, FESC, Department of Cardiology
AK Altona, Paul-Ehrlich Str. 1, D-22763 Hamburg, Germany. Email: docbergmann@mac.com

Back to November 2024 Table of Contents