Between 05/2017 and 03/2021, 9 consecutive patients received transfemoral (TF) TAVI using the Acurate neo and neo2 THV for pure non-calcified AR following an adjusted valve sizing algorithm. Assessment of prohibitive risk for SAVR [3/9 patients with previous cardiac surgery, 4/9 due to age and comorbidities, 1/9 due to previous left ventricular assist device (LVAD) implantation and 1/9 due to malignant disease] and allocation to TAVI followed current international recommendations (4, 5) after consensus of the local dedicated heart team. Written informed consent was obtained from all patients. The study was approved by the local institutional review board.

The preprocedural diagnostic work-up followed institutional standards and was described before (15): By routine, all patients received preoperative transthoracic and/or transesophageal echocardiography (TEE), a contrast-enhanced, electrocardiogram-gated MSCT for calculation for native aortic annulus dimensions and determination of adequate THV size as well as assessment of aortic root anatomy and morphology with the 3mensio Medical Imaging Software (3mensio, Medical Imaging, Bilthoven, Netherlands). Valve sizing followed an adjusted algorithm proposed by Kim et al. for aortic valve stenosis (16) with additional oversizing equivalent to 10.7 ± 2.7% of the THV diameter when compared to the native annulus diameter in this series (see Table 1). As a crude measure, an effective perimeter-derived annular diameter of 25.5 mm was considered the absolute technical maximum even though secondary measures such as annular eccentricity and diameter of the left ventricular outflow tract were also considered and may have contraindicated procedures below this value.

First line approach for all procedures was local anesthesia and/or analgosedation. All procedures were performed in a specially equipped hybrid operating suite by a dedicated team of cardiologists, cardiac surgeons and anesthesiologists. The first step of THV deployment was conducted on the beating heart using fast pacing while, during the second deployment step, rapid ventricular pacing (RVP) was used to ensure stable THV positioning. In the LVAD patient, the guidewire was placed adjacent to the LVAD inflow by fluoroscopy and TEE guidance and LVAD flow was minimized during THV deployment to avoid ventricular migration. THV function was assessed by invasive measurements of hemodynamics, aortic root angiography, and TTE.

The Boston Scientific Acurate neo (Boston Scientific, Marlborough, MA, USA) THV has a SE nitinol frame carrying porcine pericardial leaflets in a supraannular position (Figure 1). The most important difference to other self-expanding platforms is the top-down deployment with minimal protrusion of the stent toward the left ventricular outflow tract. In addition, supraannular leaflet function provides very low gradients even in small anatomies (17) and the pericardial skirt in the new generation neo 2 design seals effectively against PVL. The transfemoral delivery system has a 18Fr outer diameter shaft. In detail, first the upper crown is opened, which guarantees stable positioning and supraannular anchoring of the valve. Then the flexible stabilization arches are opened, responsible for the self-aligning properties of the valve, thereby ensuring coaxial alignment. Finally, in step two the lower crown is deployed anchoring the device inside the annulus. The THV is available in three sizes for aortic annulus sizes from 20.0 to 26.3 mm and are labeled small (S: for aortic annulus sizes 20.0–22.4 mm), medium (M: for aortic annulus sizes 22.5–24.3 mm) and large (L: for aortic annulus sizes 24.4–26.3 mm) (16, 18) in AS patients.

Baseline, intraprocedural and acute follow-up data up to 30 days were retrospectively collected and entered into a standardized database and analyzed. Clinical endpoints were adjudicated in accordance with the updated standardized VARC-2 definitions (19). Data are presented as absolute numbers and percentages for categorical variables and mean values and standard deviation for continuous variables.

All 9 consecutive patients (88.8% female, 74.4 ± 7.1 years) demonstrated an increased risk profile as reflected by common risk stratification tools (EuroSCORE II 5.5 ± 3.6%, STS PROM 6.2 ± 3.0%). Patients presented with a high comorbidity burden, including 3/9 (33.3%) with previous sternotomy, 3/9 (33.3%) patients with concomitant coronary artery disease, and 2/9 (22.2%) patients with diabetes mellitus. 7/9 patients (77.7%) of the herein investigated patients were highly symptomatic with a New York Heart Association functional class ≥III. Baseline left ventricular ejection fraction was preserved in 44.4% (4/9), moderately reduced in 33.3% (3/9), and severely reduced in 22.2% (2/9). One patient suffered from end-stage heart failure and s/p left ventricular assist device (LVAD) implantation. Detailed patient demographics are summarized in Table 2.

Procedure time, fluoroscopy time, and volume of contrast agent used were 70.9 ± 32.4 min, 23.2 ± 13.8 min, and 212.3 ± 105.8 ml, respectively. The initial positioning of the THV differed from the standard procedure for AS. The THV was positioned ~1 mm higher in the aortic annulus compared to interventions for AS. For final deployment (step two), rapid pacing was used for stable THV release. For implantation in the LVAD patient, LVAD flow was reduced prior to valve deployment.

The majority of the patients was treated under local anesthesia and/or analgosedation (8/9, 88.8%). Detailed periprocedural data are summarized in Table 3.

All-cause 30-day mortality was 0% (0/9). Device success and early safety were 100% (0/9) and 77.7% (7/9), the latter due to two cases of acute kidney. The VARC-2 adjudicated clinical endpoints—stroke, myocardial infarction, or access site complication—did not occur. No postprocedural conduction disturbance or PPM implantation was observed. Intensive care unit and hospital stay were 1.7 ± 1.1 and 12.9 ± 8.8 days, respectively. Prolonged hospital stay was due to extensive preoperative diagnostic work-up because of planned off-label procedures. Echocardiography at 30 days revealed transvalvular peak/mean pressure gradients of 15.3 ± 12.3/7.2 ± 5.5 mmHg. PVL was ≤ trace in eight patients (77.7%, 7/9) and mild in two patients (22.2%, 2/9).

For detailed outcome parameter see Table 4.

Main findings of the herein conducted series using the Acurate neo or neo2 THV in TF TAVI for treatment of pure non-calcified AR are (I) the utilized device presents encouraging results in treatment of AR in patients not eligible for surgery, (II) clinical and echocardiographic results with no documented device migration/embolization and no PVL > mild suggest advantages of the particular design features of the THV for treatment of AR and (III) device success rate suggests that the herein recommended valve sizing algorithm may be especially appropriate for the Acurate neo 2 valve for this special subset of patients.

Over the last years, an increasing number of TAVI procedures for treatment of AR has been documented. Since prevalence of AR increases with age and physicians gather more experience with interventional treatment of this specific subset of patients, this trend is expected to continue. However, TAVI for AR represents certain pre- and intraprocedural challenges which are reflected by documented learning curves and uncertainty regarding adequate valve selection and sizing algorithms (20). For the most frequently used THV systems for treatment of AR varying clinical outcomes and valve sizing algorithms are described. A review of 31 published manuscripts by Yousef et al. (21) evaluating a variety of THV (CoreValve, JenaValve, Direct Flow, Acurate TA, J-Valve, Sapien, Lotus) showed unfavorable clinical outcomes with high 30-day mortality (9.6%), high intraprocedural need for a second THV (11.3%), high PPM rate (10.7%), and a high incidence of more than moderate PVL (17.7%). Although different valve types were combined in this review, oversizing was ≤ 10% in two thirds (66.4%) of cases (21). However, with latest generation devices and increased experience improved outcomes were documented over the last few years. Registry data shows that the most frequently used THV for treatment of AR is the SE CoreValve Evolut/EvolutR (14). Although documented outcomes with this THV generation are significantly improved compared to the initial CoreValve system, mortality, PPM, and residual significant PVL rates are still remarkable with 9, 20, and 6.2% respectively (12). A series of 24 AR patients who received the Acurate neo THV achieved a device success rate of 87.5%, with moderate PVL in two patients, both when valve oversizing was >10%. Furthermore, mortality was 4.1% and PPM rate 21.1% (13). These reports and our herein described data, although comprising only a small series of patients, suggest that especially the Acurate neo2 THV may be particularly suitable for treatment of AR. Nevertheless, those cited previous reports presented significant higher rates of residual PVL and need for a second THV as compared to implantation of this THV for treatment of AS. Reasons for absence of valve migration/embolization and no PVL > mild in our series may be a combination of a modified sizing algorithm with an oversizing ratio of >10% in the majority of patients, the positioning ~1 mm higher in the aortic annulus than in interventions for AS, the x-shaped stent frame preventing distal or proximal migration as well as extensive user experience with this type of THV. However, it has to be emphasized that the Acurate neo THV presented unfavorable results regarding residual significant PVL compared to a latest generation BE THV and inferiority regarding mortality and residual significant PVL compared to a latest generation SE THV in randomized controlled trials investigating TAVI in patients with AS (22, 23). Why these outcomes differ from the herein seen hemodynamic results remains speculative but may be founded in the valve sizing algorithm and the high implantation height. Reasons for the two cases of postoperative acute kidney injury remain speculative, but are highly likely attributable to preoperative existing reduced kidney function as reflected by preoperative creatinine values and preoperative reduced left ventricular function in the majority of the herein investigated patients.

Of note, in this AR patient cohort, the second deployment step was conducted under RVP to avoid THV dislocation. In one patient with status post LVAD implantation, LVAD flow was briefly reduced to a minimum during device deployment as an additional measure to avoid proximal embolization. The herein described more pronounced oversizing algorithm did not lead to conduction disturbances or postprocedural PPM implantation. This suggests that the documented high PPM rates with other THV systems may rather result from deep implantation height rather than oversizing alone. However, current THV systems are still used off-label and anchoring in non-calcified aortic annuli carries a certain risk for device migration. Therefore, for ideal results and patient safety, THV systems with modified anchoring mechanisms, like the only recently approved JenaValve with clipping of aortic valve cusps, might be advantageous (10).

These results will have to be confirmed in larger patient numbers for further clinical evaluation.

Limitations are inherent in a single-center study design with limited patient numbers: patients were not randomized to a specific treatment or THV, therefore patient preselection with hidden confounders may apply. Furthermore, this is a purely descriptive study and conclusions regarding feasibility and especially regarding long-term safety should be drawn with caution.