Valve-in-Valve Implantation of Medtronic CoreValve Prosthesis in Patients With Failing Bioprosthetic Aortic Valves: Mid-term Outcomes From the Italian CoreValve Clinical Service Project

Objectives. We evaluated the acute and two-year safety and efficacy of using the Corevalve, Evolut R, and Evolut PRO valves for treating failed surgical bioprosthesis from the Italian CoreValve Clinical Service Project. Background. Valve-in-valve (ViV) TAVR is an emerging treatment option for failed surgical bioprosthesis. Choice of transcatheter valve is an important determinant of procedural and clinical outcomes, however, longer-term data are lacking. Methods. The Clinical Service Project is a national clinical data repository evaluating the use of implantable devices across Italy. The present multi-center analysis includes consecutive patients who underwent ViV-TAVR with the Medtronic CoreValve series between October 2008 to June 2019. Evaluated endpoints included rates of overall mortality, cardiovascular mortality, myocardial infarction, and cerebrovascular accidents at 2-year follow-up. Procedural success, complications, and echocardiographic outcomes were reported according to VARC-2 criteria. Results. A total of 139 patients (mean age, 80 ± 7 years; 47.5% male; mean STS score, 10.0 ± 9.7%) underwent ViV-TAVR with CoreValve (28.5%), Evolut R (68.6%), and Evolut Pro (2.9%) valves. Device success was achieved in 68% and acute coronary obstruction requiring PCI was observed in 4 patients (2.9%). Moderate PVL was observed in 3.7% and 2.8% of patients at 30-day and 2-year follow-up and moderate structural valve degeneration seen only 5 patients (3.6%). All-cause and cardiovascular mortality were 3.6% and 2.9% at 30 days, respectively, and 20.6% and 10.2% at 2-year follow-up. Conclusions. This real-world nationwide analysis demonstrates the acute and longer-term safety and efficacy of using the self-expanding Medtronic THV for ViV-TAVR.

J INVASIVE CARDIOL 2022;34(2):E73-E79.

Key words: perivalvular regurgitation, snorkel stenting, transcatheter aortic valve replacement, valve-in-valve

Valve-in-valve (ViV) transcatheter aortic valve replacement (TAVR) is considered a valid therapeutic option for patients with failure of a previously implanted bioprosthetic surgical aortic valve (SAV) who are deemed at high risk for re-do surgery.¹ Given the increased use of bioprosthetic surgical valves, which avoids the need for long-term anticoagulation, the potential number of ViV procedures is set to increase.² However, limited data exist about the mid- and long-term follow-up of transcatheter heart valves (THVs) in failed bioprosthesis. Previous reports from the Placement of Aortic Transcatheter Valves (PARTNER) 2 study described favorable outcomes with balloon-expandable valves at 3 years follow-up.³ However, in comparison, limited data exist regarding the safety and efficacy of using supra-annular self-expanding THVs in failed SAV. Prior data from the Valve In VAlve (VIVA) registry demonstrated good clinical outcomes in patients treated with the self-expanding THV CoreValve or Evolut R (Medtronic) at 1-year follow-up.⁴ Nevertheless, longer-term follow-up data are scarce. The aim of this analysis from the Italian Clinical Service Project focusing on ViV for failed SAV is to evaluate the 2-year safety and performance of the self-expandable CoreValve, Evolut R, and Evolut Pro THVs in a consecutive cohort of patients.

The Clinical Service Project is a national clinical data repository and medical care quality improvement project designed to evaluate and improve the use of implantable devices in Italian clinical practice (clinicaltrials.gov NCT01007474). The key scientific aims of this Clinical Service Project are to (1) describe the use of implantable devices at each institution; (2) identify predictors or clinical variables correlated with clinical outcomes; and (3) evaluate and test methods to optimize patient therapy. This study was approved by each site’s institutional review board or medical director and conforms to the principles outlined in the Declaration of Helsinki. For each patient, the indication for a ViV procedure was based on local consensus of a multidisciplinary heart team and all patients signed informed, written consent. Consecutive patients undergoing TAVI with the CoreValve system/Evolut R/Evolut Pro for a failed surgical bioprosthesis were deemed eligible for the purpose of this analysis.

Procedure and follow-up. The THVs used for the ViV procedure have already been described elsewhere.⁵ THV sizing was based on preprocedural multislice computed tomography (MSCT) analysis when available or by consulting sizing charts of the surgical prosthesis. Conventional transfemoral access was the default route. However, in cases with challenging transfemoral routes, choice of alternative access was left to the operator’s discretion, with a preference for trans-subclavian access. In cases where subclavian access was required, general anesthesia or local anesthesia with deep sedation was employed in accordance with each institution’s protocol. The acute performance of the implanted prosthesis was assessed interprocedurally using echocardiography. In the presence of residual high gradients or significant paravalvular regurgitation (PVL), postdilation was performed according to the operator’s choice. Balloon-valve fracture was not performed amongst the reported cases. Choice and modality of coronary protection was left to the operator’s discretion based on preprocedural MSCT analysis of high-risk anatomical features as previously described.^6,7

Before discharge, patients underwent echocardiography to assess the final result of the THV implantation in the failed bioprosthesis. In addition, patients were encouraged to return to the outpatient clinic for follow-up echocardiography at 30 days, 6 months and then annually to determine the performance and durability of the implanted valve. Echocardiographic outcomes were evaluated according to the Valve Academic Research Consortium (VARC) 2 criteria.⁸

Endpoints. All the endpoints were defined and parameters analyzed according to the VARC-2 definitions.⁸ The primary endpoint was the combined rate of overall mortality, cardiovascular mortality, myocardial infarction, and cerebrovascular accidents at 2 year follow-up. Device success was also defined by VARC-2 criteria as absence of procedural mortality, correct positioning of a single prosthetic heart valve and intended performance of the prosthetic heart valve (no patient-prosthesis mismatch (PPM) and mean aortic valve gradient <20 mm Hg or peak velocity <3 m/s and no moderate or severe prosthetic valve regurgitation). At follow-up, patients were also evaluated for the occurrence of structural valve degeneration (SVD) that was defined according to standardized proposed definitions.⁹

Statistical analysis. Descriptive statistics have been used to summarize patient characteristics. Categorical variables are reported as number (percent) and continuous variables as mean ± standard deviation or median (25th–75th interquartile range), depending on variable distribution. Time-to-event curves using the Kaplan-Meier method were used to assess the events occurring at follow-up. We evaluated variations in mean transaortic gradient over time using Generalized Estimating Equations (GEEs) models for continuous outcomes using the patient as the subject and follow-up visits as the repeated factor. Statistical analyses were performed using the SAS software, version 9.4 (SAS Institute).

The main baseline demographic and clinical features of the study population are listed in Table 1. Mean age was 80 ± 7 years, approximately one-half of the patients (47.5%) were male, and the mean Society of Thoracic Surgeons score was 10.0 ± 9.7%. The main mechanism of prosthetic failure was valve regurgitation (n = 79, 59.4%) and the mean transvalvular gradient was 43.7 ± 14 mm Hg. Concomitant coronary artery disease was present in 48 patients (35%). Evolut R was the most widely used type of prosthesis (68.6%) and in the majority of the cases (62.5%) a 23 mm THV was implanted. Postdilation was required in one-third of patients (35.3%).

Procedural characteristics are listed in Table 2 and study outcomes are presented in Table 3. VARC-2 defined device success was achieved in 68% of the patients. Five procedures failed because of the need for a second valve implant. Four patients experienced an abrupt occlusion of the left main coronary artery (LMCA), which required an urgent percutaneous coronary intervention (PCI) with implantation of a drug-eluting stent (DES). An additional patient underwent LMCA PCI, using the “chimney” technique, as they were deemed to be at high risk for delayed coronary obstruction. In all 5 cases the PCI was successfully performed without further events during the index hospitalization. The overall in-hospital mortality was favorable with a mortality rate of 1.4% and a need for pacemaker implant of 12.1% (n = 13). The rate of in-hospital and 30-day cerebrovascular events was 2.2% and 2.9%, respectively.

Echocardiographic outcome and follow-up. At 30-days, the vast majority of patients had none/trace or mild PVL (Figure 1), with moderate regurgitation present in 3.7% of the ViV cohort. No patients had severe regurgitation. The proportion of patients with moderate PVL remained low at 2-year follow-up (2.8%). Similarly, the THV performance based on the transvalvular gradients was good and the mean gradient remained stable over the 2-year follow-up (Figure 2). Five patients had evidence of moderate SVD, all manifesting with an increase in the transvalvular mean gradient (mean increase 10.6 ± 6.1 mm Hg).

Clinical follow-up. The 2-year clinical follow-up are reported in Figure 3. The Kaplan-Meier (KM) estimate for overall and cardiovascular mortality was 20.6% and 10.2%, respectively. The overall KM estimate for myocardial infarction and cerebrovascular events was 4.5% and 4.3% respectively with a higher proportion of events occurring during the index hospitalization.

The main findings of our report are as follows:

1. Use of the CoreValve/Evolut R or the Pro prosthesis for a ViV procedure was safe and associated with an acceptable device success rate.
2. The rate of acute coronary obstruction was not negligible (n = 4; 2.8%) and required urgent PCI of the left main during the procedure. In another case, deemed high risk for delayed coronary obstruction, LMCA PCI was planned and performed prior to THV implantation.
3. At 2-year follow-up the echocardiographic outcome was favorable with mean gradients remaining stable and with 97.2% of the population having less than moderate PVL.
4. Two-year clinical follow-up demonstrated a high-rate of mortality (20.6%) in this cohort of patients with half occurring for cardiovascular reasons.

ViV-TAVR is an emerging less invasive treatment strategy for degenerated aortic surgical bioprosthesis. In these patients, open-heart redo surgery is limited by a number of complications related to re-sternotomy and the need for cardiopulmonary bypass. Re-sternotomy is associated with a longer procedural duration due to the presence of adhesions and is associated with higher rates of morbidity and mortality compared to first-time surgery. Existing data evaluating redo aortic valve surgery are conflicting. In a large series of 155 patients with prohibitive surgical risk (mean Logistic EuroScore,²⁷ rate of endocarditis 27.1%), the in-hospital mortality was 5.8%.¹⁰ In more selected patients undergoing redo surgery with the sutureless Perceval S aortic bioprosthesis (Sorin Biomedica Cardio Srl), the authors reported no in-hospital deaths and positive outcomes in terms of bioprosthesis performance (mean gradient 10.3 ± 1.5 mm Hg) without any significant PVL.¹¹

A prior meta-analysis which compared 498 patients undergoing ViV-TAVR or redo surgical aortic valve replacement (SAVR) found no differences in perioperative mortality (4.4% vs 5.7%, P=.83; I²=0%) but the surgical cohort had a lower risk of severe patient prosthesis mismatch (3.3% vs 13.5%; P=.03; I²=0%) and less mild or greater paravalvular leak (5.5% vs 21.1%; P=.03; I²=37%).¹² A further meta-analysis of 595 patients comparing ViV-TAVR with SAVR demonstrated comparable procedural, 30 day and 1-year mortality rates as well as similar rates of stroke, major bleeding and vascular complications. ViV-TAVR was associated with a lower risk of permanent pacemaker implantation and a slight trend towards increased paravalvular leak.¹³ More recently, a propensity-matched analysis of 558 patients found that compared to re-do surgery ViV-TAVR was associated with lower rates of permanent pacemaker implantation and blood transfusions, shorter hospital stay and a greater overall survival at 5-years follow-up (76.8% vs 66.8%).¹⁴

The success of ViV-TAVR depends upon a wide variety of parameters, most importantly the type of failed surgical prosthesis, the height of implantation as well as the type of THV, which can impact both PVL and residual gradients. Balloon-expandable valves (Myval, Lifesciences and Sapien, Edwards Lifesciences), the self-expandable Portico (Abbott), and the mechanically expandable Lotus (Boston Scientific) valves all have an intra-annular design, making them prone to achieving elevated postprocedural residual gradients. Even if good results can be achieved with meticulous positioning of all THVs, supra-annular compared to intra-annular valves have a better hemodynamic profile with, lower postprocedural residual gradients and more favorable outcomes at long-term follow-up.^15,16 In our analysis, we observed a 30-days mean gradient of 15.8 mm Hg which is consistent with previous reports of the Medtronic CoreValve series for failed biological prosthesis.^17,18 A possible explanation for the elevated postprocedural gradients observed in our cohort could be the increased use of small-sized THVs (62.5% of the patients received a 23 mm THV). Although, post-dilatation to further optimize the result was performed in 33.8% of cases, no patients underwent balloon valve fracture, which could have potentially further improved the postprocedural hemodynamics. Nevertheless, compared to balloon-expandable valves, the supra-annular Medtronic THVs have demonstrated superior performance in terms of residual gradients as reported by Bleiziffer et al in their recent registry evaluating the 8-year outcomes of a large cohort of ViV procedures. Moreover, Medtronic self-expandable valves are less prone to reinterventions at follow-up, a relevant factor for this high-risk cohort.¹⁶

Our data confirm the satisfactory results observed following ViV-TAVR, with stability of gradients, a low percentage of moderate SVD and no cases of severe SVD observed after 2-year follow-up. Moreover, no patients required reintervention during follow-up. These data are in line with previous reports of the Medtronic self-expandable valves. In particular, the CoreValve US Expanded Use Study reported an overall rate of SVD of 11.0% at 3 years (2.7% of patients developed severe SVD and 8.2% of patients developed moderate SVD). The rate of valve reintervention over the 3-year follow-up period was <5%.¹⁸ Furthermore, they observed moderate PVL in 3.4% and 4.5% of the population at 30-day and 2-year follow-up. In contrast, we observed moderate PVL in 3.7% at 30-days and 2.8% at 2-year follow-up. This apparent disparity could be explained by the use of newer generation Evolut R and Pro valves, which offer a better sealing and adaptation to the failed bio-prosthesis compared to the first generation CoreValve.¹⁹ However, in comparison, use of balloon-expandable valves for ViV-TAVR was associated with less moderate PVL.¹⁶ In the VIVID registry, which included different types of Edwards-Sapien THV, the reported rate of moderate to severe PVL was 2.7% and a similar rate (2.5%) was reported when using the latest generation of Sapien 3 valves.³

All ViV-TAVR procedures were successfully completed and no intraprocedural deaths were reported. Interestingly, we reported four cases in which an impairment of coronary flow was observed and emergent PCI was performed using the chimney/snorkel technique to avoid the risk of complete coronary occlusion.²⁰ In another case, coronary protection in the way of prophylactic LMCA stenting was performed due to the perceived elevated risk for delayed coronary obstruction. The overall rate of patients who required concomitant coronary intervention was not negligible (3.6%). This was slightly higher compared to the report by Ribeiro et al where 37 out of 1612 (prevalence 2.3%) patients developed coronary obstruction.²¹ Among these 37 patients, approximately half of the patients (48.7%) died at 30 days as a consequence of the myocardial injury related to the coronary obstruction. In our population, all the patients who experienced a coronary obstruction survived and no significant cardiovascular events were observed during the follow-up. It is possible that the positive outcome observed in our population is related to the timing of occlusion. Except for one patient, who experienced signs of coronary occlusion once in the intensive care unit, all the other patients had EKG modifications immediately after valve deployment. In all the cases successful PCI and stent deployment was performed with prompt restoration of blood flow without incurring significant myocardial damage. This was despite the recognized challenge of re-engaging coronary arteries following CoreValve/Evolut implantation in native valve anatomies.^22-24 This challenge can be further amplified due to misalignment of the THV and surgical commissures.²⁵ Recently, refinements in implantation techniques have been reported which may improve the commissural alignment using the Core-valve/Evolut valves, however these have not been evaluated in the setting of ViV-TAVR.²⁵

Study limitations. The present research has several limitations. First of all, the events reported during the index procedure and at follow-up were not evaluated by an external steering committee but were adjudicated by local investigators. Moreover, echocardiographic follow-up was performed without a centralized core-lab and at the discretion of the operator. Most of the patients did not undergo routine follow-up with serial echocardiographic evaluations, which makes our data on SVD and trans-prosthetic gradients more reflective of real-world findings. Moreover, the procedures were performed over a large timeframe using three different generations of the Medtronic THV which could have influenced the final results. Finally, no cases of bioprosthetic valve ring fracture were included in our cohort, however the long-term clinical outcomes of patients who undergo ring fracture are still unknown.

Our analysis from the Italian CoreValve Clinical Service Project demonstrates that the self-expanding Medtronic THV remains safe, durable, and effective over 2-years follow-up for the indication of failed surgical valve in patients unsuitable for re-do surgical aortic valve replacement. We observed a low rate of severe SVD and no cases of reinterventions. Patients who developed acute coronary obstruction were successfully managed.

• Transcatheter Tricuspid Valve-in-Valve Replacement Via Right Internal Jugular Vein Access in a 39-Year-Old Woman
• Transcatheter Mitral Valve-in-Valve and Valve-in-Ring Replacement in High-Risk Surgical Patients: Feasibility, Safety, and Longitudinal Outcomes in a Single-Center Experience
• Percutaneous Closure of a Poorly Tolerated Ventricular Septal Defect After Triple Valvular Prosthesis Surgery
• Percutaneous Mitral Valve Repair With MitraClip in Inoperable Patients With Severe Mitral Regurgitation Complicated by Cardiogenic Shock