Circulation Journal
Online ISSN : 1347-4820
Print ISSN : 1346-9843
ISSN-L : 1346-9843

This article has now been updated. Please use the final version.

Transcatheter Aortic Valve Implantation (TAVI) for Native Aortic Valve Regurgitation　― A Systematic Review ―
Altayyeb Yousef, Zachary MacDonald, Trevor Simard, Juan J. Russo, Joshua Feder, Michael V. Froeschl, Alexander Dick, Christopher Glover, Ian G. Burwash, Azeem Latib, Josep Rodés-Cabau, Marino Labinaz, Benjamin Hibbert
Author information
Keywords: Aortic insufficiency, Aortic valve regurgitation, Transcatheter aortic valve implantation, Transcatheter aortic valve replacement
JOURNAL FREE ACCESS FULL-TEXT HTML Advance online publication
Supplementary material

Article ID: CJ-17-0672

The final version of this article is now available: Vol. 82 (2018), No. 3 pp. 895-902
Details
Download PDF (408K)
Download citation RIS

(compatible with EndNote, Reference Manager, ProCite, RefWorks)

BIB TEX

(compatible with BibDesk, LaTeX)

Text
How to download citation
Contact us
Abstract

Background: Transcatheter aortic valve implantation (TAVI) has become the standard of care for management of high-risk patients with aortic stenosis. Limited data is available regarding the performance of TAVI in patients with native aortic valve regurgitation (NAVR).

Methods and Results: We performed a systematic review from 2002 to 2016. The primary outcome was device success as per VARC-2 criteria. Secondary endpoints included procedural complications, and 30-day and 1-year mortality rates. A total of 175 patients were included from 31 studies. Device success was reported in 86.3% of patients – with device failure driven by moderate aortic regurgitation (AR ≥3+) and/or need for a second device. Procedural complications were rare, with no procedural deaths, myocardial infarctions or annular ruptures reported. Procedural safety was acceptable with a low 30-day incidence of stroke (1.5%). The 30-day and 1-year overall mortality rates were 9.6% and 20.0% (cardiovascular death, 3.8% and 10.1%, respectively). Patients receiving 2nd-generation valves demonstrated similar safety profiles with greater device success compared with 1st-generation valves (96.2% vs. 78.4%). This was driven by the higher incidence of second-valve implantation (23.4% vs. 1.7%) and significant paravalvular leak (8.3% vs. 0.0%).

Conclusions: TAVI demonstrates acceptable safety and efficacy in high-risk patients with severe NAVR. Second-generation valves may afford a similar safety profile with improved device success. Dedicated studies are needed to definitively establish the efficacy of TAVI in this population.

Since 2002, transcatheter aortic valve implantation (TAVI) has evolved into the standard of care for a number of patients with native valve aortic stenosis (AS).1 Specifically, TAVI has proved efficacious in patients with prohibitive operative risk2 or high operative risk, when compared with medical or surgical therapy.3 Based on this early success, operators have expanded the scope of eligible TAVI candidates to an ever growing list of aortic valve pathologies.4 Indeed, TAVI is often used in patients with low-flow low-gradient severe AS,5 stenotic bicuspid aortic valves (BAVs),6 and in patients with failing bioprostheses.7 Recently, increasing numbers of reports of TAVI being used in patients with aortic regurgitation (AR) in native aortic valves have been published.

Patients with multiple comorbidities suffering from severe native aortic valve regurgitation (NAVR) with elevated surgical risk present a challenging dilemma to cardiologists. Akin to prohibitive or high operative risk patients with AS, this cohort of patients is rapidly becoming the focus of TAVI operators as our experience and expertise with TAVI grows. Repairing NAVR with TAVI would seem to be a natural progression of work in the AS population. However, the lack of a calcified native valve apparatus to anchor the TAVI valve in place is an important anatomic difference necessitating unique procedural approaches and device performance. To date, there is no clinical trial evaluating the outcomes of TAVI in NAVR to guide clinical decisions. Accordingly, we conducted a systematic review of all reported cases to evaluate the contemporary performance of TAVI in patients with isolated NAVR.

Methods

Following the PRISMA guidelines checklist (Supplementary File 1),8 all abstracts published between 1 January 2002 and 12 February 2016 in Medline, Embase, Pubmed, Cochrane Controlled Trials Register and Cochrane Database of Systematic Reviews were systematically reviewed. The following medical search terms were used: aortic valve insufficiency, aortic regurgitation, regurgitant aortic valve, aortic incompetency, incompetent aortic valve, NAVR, non-calcific aortic valve, transcatheter aortic valve replacement, transcatheter aortic valve or transfemoral aortic valve or transaortic aortic valve or transapical aortic valve or transcutaneous aortic valve or percutaneous aortic valve or TAVI or TAVR. Additional studies were identified by reviewing the references and citing articles of identified manuscripts.

Inclusion criteria included TAVI as the reported or intended therapeutic intervention, and isolated severe AR. Exclusion criteria included: (1) mixed aortic valve disease (any combined degree of AS and AR) (2) prior intervention to the aortic valve (prosthetic or repaired aortic valve), (3) concomitant procedure(s) at the time of TAVI and (4) non-English language publication. To prevent repeat reporting of patient cohorts, whenever a co-author was identified between abstracts, the most recent publication was included unless the patient population was clearly distinct between studies.

Manuscripts were identified by reviewing the titles and abstracts of studies with selection performed by 2 coauthors and conflicts resolved by consensus. Two coauthors independently reviewed the full manuscripts and performed data extraction, with conflicts resolved by consensus.

The primary endpoint of the study was device success, defined according to the Valve Academy Research Consortium (VARC-2) consensus document.9 Specifically, device success was defined as a composite endpoint of (1) absence of procedural death AND (2) correct positioning of a single prosthetic heart valve into the proper anatomic location AND (3) intended performance of the prosthetic heart valve with no significant AR (defined as moderate (AR 3+) or severe (AR 4+) AR. Thus, where the VARC criteria were not explicitly stated, redefinition of device success reported in the included studies was standardized to these criteria.

Secondary endpoints included incidence of mild to moderate AR or worse (AR ≥2+), valve malposition (defined as valve embolization, migration or ectopic deployment) at any time, need for a second TAVI, stroke, major bleed, annulus rupture, myocardial infarction and pacemaker insertion.

All continuous variables are described as means±standard deviation for parametric variables and medians with interquartile range (IQR) for non-parametric variables. Categorical variables are described as number (%). Unadjusted analyses of the primary and secondary outcomes were made using Chi-square or Fisher’s test as appropriate. Analysis was performed using Sigmaplot v.13.0 (Systat Software, Chicago, IL, USA).

Results

We identified the population, intervention and outcomes of interest a priori. Specifically, we wished to evaluate the clinical performance of TAVI in patients with NAVR. Our search strategy identified 4,422 candidate studies of which 4,352 were excluded after brief review (Figure). In total, 70 publications underwent systematic review and 39 further studies were excluded (Supplementary File). In total, 31 publications with a total of 175 patients were included in our analysis. Articles were published between 2010 and 2016; there were 6 case series/cohort studies,10–15 20 single patient case reports,16–36 and 5 online abstract studies (Table 1).37–41 The characteristics of the excluded studies are shown in Table S1, and the appropriate quality and risk of bias assessments Table S2 and Table S3, respectively.

Figure.

Selection of studies for review of transcatheter aortic valve implantation for native aortic valve regurgitation.

Table 1. TAVI for NAVR: Characteristics of Included Studies
Author
(year)		 Country Study
design		 No. of
patients		 Age
(years)		 Male STS
score		 Logistic
Euroscore		 Reason
surgery
declined		 Valve
type		 VARC
criteria
reported		 Follow-up
Roy et al
(2013)10		 Worldwide/
multicenter		 Case
series		 42 75.3±8.8 20 10.2±5.3 NR High risk CoreValve Yes 12 months
Seiffert et al
(2014)12		 Germany/
multicenter		 Case
series		 31 73.8±9.1 20 5.4±3.6 23.6±14.5 Inoperable JenaValve Yes 6 months
Testa et al
(2014)11		 Italy/
multicenter		 Case
series		 26 73±10 16 13.1±2 24±8 Inoperable CoreValve Yes 12 months
Schofer et al
(2015)13		 Europe/
multicenter		 Case
series		 11 74.7±12.9 4 8.84±8.90 19.9±7.1 High risk Direct Flow Yes 30 days
Wendt et al
(2014)14		 Germany Case
series		 8 72.5±8.4 5 7.9±3.4 34.0±7.9 High risk Acurate TA Yes 12 months
Wei et al
(2015)15		 China Case
series		 5 74.8±8.9 3 NR 29.59 High risk J-Valve No 6 months
Segev et al
(2013)34		 Israel Case
report		 2 63; 86 2 NR NR NR CoreValve No 30 days
Ducrocq et al
(2010)16		 France Case
report		 1 36 0 NR NR High risk CoreValve No NR
Krumsdorf et al
(2011)17		 Germany Case
report		 1 87 0 22 46 Inoperable CoreValve No 6 weeks
D’Ancona et al
(2012)18		 Germany Case
report		 1 63 1 NR NR NR Edwards No NR
Santini et al
(2012)36		 Italy Case
report		 1 53 1 NR NR NR CoreValve No NR
Yeow et al
(2012)20		 Australia Case
report		 1 74 0 5.1 21.2 High risk CoreValve No 6 months
Hildebrandt
et al (2013)21		 Germany Case
report		 1 75 1 28 NR High risk CoreValve No NR
Zanuttini et al
(2013)22		 Italy Case
report		 1 75 1 NR 36 High risk CoreValve No NR
Chiam et al
(2014)23		 Singapore Case
report		 1 43 1 NR NR Inoperable CoreValve No 6 months
Cholteesupachai
et al (2014)24		 Denmark Case
report		 1 87 1 7.4 NR Inoperable CoreValve No 3 months
Kiefer et al
(2014)25		 Germany Case
report		 1 72 0 7.2 NR NR CoreValve No No
follow-up
Krause et al
(2014)26		 Germany Case
report		 1 65 0 NR NR NR CoreValve No 12 months
Maureira et al
(2014)27		 France Case
report		 1 68 1 5.43 11.05 NR CoreValve No 6 months
Singh et al
(2014)35		 USA Case
report		 1 72 1 12.4 NR High risk Edwards No 6 months
Cabasa et al
(2016)48		 USA Case
report		 1 64 1 NR NR High risk CoreValve No 4 months
Hebbar et al
(2015)31		 France Case
report		 1 56 0 NR NR High risk CoreValve No 3.5 years
Kornberger
et al (2015)33		 Germany Case
report		 1 64 1 15 NR High risk Edwards No 14 days
Özpelit et al
(2015)32		 Turkey Case
report		 1 85 1 NR NR High risk CoreValve No 6 weeks
Wöhrle et al
(2015)30		 Germany Case
report		 1 78 0 NR NR High risk Lotus Yes NR
Cerillo et al
(2016)28		 Italy Case
report		 1 87 0 7.3 20.9 High risk Acurate Neo No 3 months
Guo et al
(2015)41		 China Abstract 18 73.8±3.7 NR NR 24.1±4.5 High risk J-Valve No 30 days
Munoz-Garcia
et al (2015)39		 Spain Abstract 10 79.2±4.9 NR NR 15.3±8 High risk CoreValve No Up to
1 year
Pacchioni et al
(2011)40		 Italy Abstract 1 63 0 NR NR Inoperable Edwards &
CoreValve		 No 1 year
Lavee et al
(2013)37		 Israel Abstract 1 55 0 NR NR High risk CoreValve No 3 months
Ng et al
(2013)38		 Singapore Abstract 1 49 1 NR NR High risk CoreValve No NR

NAVR, native aortic valve regurgitation; NR, not reported; TAVI, transcatheter aortic valve implantation.

Baseline characteristics are summarized in Table 2. Briefly, the median age of the population studied was 73.8 year old (IQR 73.0–75.3) with 55.7% of the population being male. As expected, the majority (93.5%) of patients had severe symptoms (NYHA class III/IV) and 65 patients (49.6%) had established coronary artery disease. The mean STS score and Log-Euroscore were 9.5% and 23.8% respectively. Most patients had severe AR (4+); however, 26.9% patients reported had only moderate AR (3+). Only 8 studies reported quantitative assessment of AR severity and 3 publications did not report the details of AR severity. The etiology of AR was noted for 141 patients (80.6%), with degenerative aortic valve disease being the predominant etiology. Although no studies specifically stated an absence of BAV features, it appears unlikely any were included in the analysis. Dilated aortic root accounted for only 15.0% of reported cases.

Table 2. Baseline Patient Characteristics
  No. (% of available data)
n=175
Demographics
　Age, years, median (IQR) 73.8 (73.0–75.3)
　Male 82 (55.7)
Comorbidities
　Diabetes mellitus 27 (20.6)
　Hypertension 71 (45.2)
　Atrial fibrillation 25 (19.1)
　NYHA Class 3 or 4 145 (93.5)
　　LVEF, median (1 st–3rd IQR) 45.5 (45.0–46.8)
　Prior cerebrovascular accident 24 (18.3)
　Peripheral artery disease 18 (13.7)
　Chronic obstructive airways disease 24 (18.3)
　Other chronic lung disease 14 (10.7)
　Chronic kidney disease 22 (22.0)
　　Serum creatinine, median (1 st–3rd IQR) 120.9 (115.0–120.9)
　Pulmonary hypertension 35 (24.5)
　Coronary artery disease 65 (49.6)
　Myocardial infarction 26 (19.4)
　Percutaneous coronary intervention 22 (16.8)
　Coronary artery bypass graft 26 (18.3)
　Thoracic aorta intervention 4 (3.1)
　Permanent pacemaker 3 (2.3)
　Anemia or prior gastrointestinal bleed 24 (18.3)
　Malignancy 6 (4.6)
　Prior history of infective endocarditis 5 (3.8)
Indication for TAVI
　STS score, mean (SD) 9.5 (3.4)
　Logistic Euroscore, mean (SD) 23.8 (4.8)
　Porcelain aorta 4 (2.3)
　Other* 15
Aortic valve characteristics
　Reported AR severity 131 (74.9)
　　3+ 35 (26.9)
　　4+ 96 (73.8)
　Aortic calcification (reported) 105 (60.0)
　　No calcification 71 (51.8)
　　Mild calcification 28 (20.4)
　　Moderate calcification 3 (2.2)
　　Severe 1 (0.7)
　　Not quantified or not reported 72 (32.4)
　Method of obtaining annulus size
　　Computed tomography scan 96 (68.1)
　　Echocardiogram 75 (53.2)
　Annulus diameter, mean (SD) 24.6 (4.2)
Mechanism of aortic regurgitation 141 (80.6)
　Degenerative aortic valve 80 (56.7)
　Dilated aortic root 21 (15.0)
　Remote infective endocarditis 15 (10.7)
　Inflammatory disease 6 (4.3)
　Radiation therapy 5 (3.6)
　Trauma 1 (0.7)
　Rheumatic 1 (0.7)
　Unknown 12 (8.6)

*Other factors not accounted for by conventional risk scores. AR, aortic regurgitation; IQR, interquartile range; LVEF, left ventricular ejection fraction; TAVI, transcatheter aortic valve implantation.

Of the 175 patients undergoing TAVI for NAVR, the majority (55.4%) received a 1st-generation self-expandable Medtronic CoreValve (Medtronic, Inc., Minneapolis, MN, USA) (Table 3); 2nd-generation valves, including the JenaValve transcatheter heart valve (JenaValve Technology GmbH, Munich, Germany), or the J-ValveTM system (JC Medical Inc., CA, USA), which were the most frequently reported 2nd-generation valves, were used in 30.8% of patients. The transfemoral approach was the most common percutaneous access site, used in 83 (50.6%) patients, and the transapical approach was used in 64 cases (39.0%); 3 studies did not report the access site used in 11 patients of the total cohort.

Table 3. TAVI for NAVR: Procedural Information
  Total no.
(% of reported)
Valve type (reported ) 176* (100)
　Sapien 3† (1.7)
　CoreValve 98 (55.4)
　Direct Flow 11 (6.2)
　JenaValve 31 (17.5)
　J-Valve 23 (13.0)
　ACURATE TA 8 (4.5)
　ACURATE neo 1 (0.6)
　Lotus 1 (0.6)
　Medtronic Engager 1 (0.6)
Approach (reported) 164 (93.7)
　Femoral 83 (50.6)
　Subclavian 9 (5.5)
　Aortic 5 (3.0)
　Carotid 2 (1.2)
　Apical 64 (39.0)
　Axillary 1 (0.6)
Size of valve (reported) 166* (94.3)
　23 6 (3.4)
　25 21 (11.9)
　26 17 (9.6)
　27 43 (24.4)
　29 57 (32.2)
　31 22 (12.4)
Oversizing (reported) 116 (65.9)
　≤10% 77 (66.4)
　10–15% 10 (8.6)
　15–20% 5 (4.3)
　≥20% 24 (20.7)

*One patient received two different valves after device embolization. †One patient received an Edward Sapien valve, which was replaced with Corevalve. Abbreviations as in Table 1.

Among patients with isolated NAVR, device success (i.e., the primary endpoint) was achieved in 151 (86.3%) (Table 4). Device failure was primarily caused by significant (moderate or severe) paravalvular leak (PVL; AR ≥3+), which occurred in 8 patients. In addition, there was a total of 6 cases of valve malposition (2 valve migrations,25,37 3 valve embolizations,17,32,41 and 1 ectopic deployment40). In addition, 3 patients required immediate surgical aortic valve replacement,11,38,41 and another patient had a late valve embolization event requiring surgical intervention 3 days post TAVI. Notably, 16 patients required a second TAVI because of device failure, for an incidence of 11.3%.11–13,26,36

Table 4. TAVI for NAVR: Combined Primary and Secondary Endpoints
Outcome n (%)
Primary outcomes
　Device success 151/175 (86.3)
　　AR ≥3+ 8/175 (4.6)
　　Valve malposition 6/175 (3.4)
　　　Valve migration 2
　　　Valve embolization 3
　　　Ectopic valve deployment 1
　　Second valve implanted 16/175 (11.3)
　　Conversion to SAVR 4/175 (2.3)
Secondary outcomes
　Procedural outcomes
　　Procedural death 0/175 (0.0)
　　Annular rupture 0/175 (0.0)
　　Post TAVI AR ≥2+ 31/175 (17.7)
　　30-day CVA 2/133 (1.5)
　　Major bleeding 16/139 (11.5)
　　Acute MI 0/139 (0.0)
　　Pacemaker 17/159 (10.7)
　　AKI 9/149 (6.0)
　Follow-up outcomes
　　30-day death 16/166 (9.6)
　　30-day cardiovascular death 6/156 (3.8)
　　NYHA III/IV at 30 days 19/105 (18.1)
　　Post TAVI LVEF, % median 46.7 (in 126 patients)
　　1-year death 16/80 (20.0)
　　1-year cardiovascular death 8/79 (10.1)

AKI, acute kidney injury; AR3+, moderate aortic regurgitation; AR2+, mild to moderate aortic regurgitation; CVA, cerebrovascular accident; LVEF, left ventricular ejection fraction; MI, myocardial infarction; NYHA, New York Heart Association class; SAVR, surgical aortic valve replacement. Other abbreviations as in Table 1.

Only 2 patients (1.5%) had a stroke by 30 days and there were no reported procedural deaths or cases of annulus rupture or myocardial infarction. However, major bleeding and pacemaker insertion occurred in 16 patients (11.5%) and 17 patients (10.7%), respectively. Acute kidney injury was described in 9 (of 149) patients (6.0%). Only 1 patient was reported to have a life-threatening bleed from pericardial tamponade without specification of the mechanism.

In the 166 patients with reported 30-day outcomes the mortality rate was 9.6%, but only 6 patients died of cardiovascular cause (3.8%). Notably, TAVI resulted in marked improvements in functional class, with 81.9% of patients achieving NYHAI/II status. Survival data at 1 year was only available for 80 patients. Of them, 16 (20.0%) died in the first year post TAVI, with 10.1% from cardiovascular causes.

The Edwards Sapien and Medtronic CoreValve are 1st-generation TAVI valves with subsequent iterations and newer designs commonly referred to as 2nd-generation valves.42 Nearly half of the patients in the cohort underwent 2nd-generation TAVI for treatment of NAVR. Notably, outcomes with 2nd-generation valves were significantly improved over those with 1st-generation systems. Specifically, device success was 78.4% with 1st-generation TAVI vs. 96.2% with 2nd-generation TAVI (Table 5). This was predominantly driven by fewer cases of significant AR (8.3% vs. 0.0%) and fewer cases of needing a second valve implanted (23.4% vs. 1.7%).

Table 5. Comparison of 1st- and 2nd-Generation Valves for TAVI in NAVR
Outcome 1st-generation valves,
n (%)		 2nd-generation valves,
n (%)
Primary outcome
　Device success 76/97 (78.4) 75/78 (96.2)
　　AR ≥3+ 8/97 (8.3) 0/78 (0.0)
　　Valve malposition 5/97 (5.2) 1/78 (1.3)
　　　Valve migration 2/97 (2.1) 0/78 (0.0)
　　　Embolization 2/97 (2.1) 1/78 (1.3)
　　　Ectopic valve deployment 1/97 (1.0) 0/78 (0.0)
　　Second valve implanted 15/97 (23.4) 1/59 (1.7)
　　Conversion to SAVR 2/97 (2.1) 2/78 (2.6)
Secondary outcomes
　Immediate procedural outcomes
　　Procedural death 0/97 (0.0) 0/78 (0.0)
　　Annular rupture 0/97 (0.0) 0/78 (0.0)
　　Post TAVI AR ≥2+ 31/97 (32.0) 0/78 (0.0)
　　Major CVA, 30 days 2/77 (2.6) 0/56 (0.0)
　　Major bleeding 13/82 (15.9) 3/57 (5.3)
　　Acute MI 0/82 (0.0) 0/57 (0.0)
　　Pacemaker 12/83 (14.5) 5/76 (6.6)
　　AKI 7/92 (7.6) 2/57 (3.5)
　Follow-up outcomes
　　30-day death 11/92 (12.0) 5/74 (6.8)
　　30-day cardiovascular death 5/82 (6.1) 1/74 (1.4)
　　NYHA III/IV at 30day 12/50 (24.0) 7/55 (12.7)
　　1-year death 16/72 (22.2) 0/8 (0.0)
　　1-year cardiovascular death 8/71 (11.3) 0/8 (0.0)

Abbreviations as in Table 4.

Discussion

AR results in significant morbidity and mortality, and in some high-risk populations NAVR is the dominant aortic lesion. As such, the number of patients with NAVR and prohibitive surgical risk will continue to rise, representing a unique clinical challenge for which TAVI may offer a novel therapeutic approach. TAVI has emerged as the preferred therapeutic option for patients with severe aortic valve stenosis and high or prohibitive operative risk. Although intuitive that this approach could be extended to patients with NAVR, large datasets are lacking. We report the outcomes in 175 patients who underwent TAVI for isolated NAVR, demonstrating that in select patients TAVI is feasible, safe and efficacious. Specifically, we report acceptable 30-day and 1-year mortality rates (9.6% and 20.0%, respectively) with a device success rate of 86.3%. Importantly, there was low reported procedural morbidity. These findings are in line with similar prohibitive-risk patients undergoing TAVI for AS,2,43 and other off-label TAVI cohorts such as patients with BAV.6,44 This supports an important role for prospective studies assessing TAVI in NAVR, while providing early guidance for heart teams considering TAVI as part of a treatment plan.

A major concern regarding TAVI for isolated NAVR is the potential for device failure, given the absence of a calcified annulus in which to anchor the device. Although early studies reported modest device success, as low as 79.1%,11 findings are likely in part related to the early experience, the generation of valve used and the variable definitions of device success. In our study, we only included patients who had moderate (3+) or severe (4+) AR and we used the VARC-2 definition of outcomes whenever possible to standardize comparisons between studies.9

NAVR presents unique challenges for TAVI as a therapy. First, patients with AR generally have a large annulus. The mean aortic annulus in this study was 24.6 cm and almost half of the patients were treated with the largest available valve sizes (29 mm and 31 mm prosthesis sizes). Hence, appropriate sizing can be limited by the availability of large enough prosthetic valves, potentially contributing to increased incidence of PVL. Second, valve calcification is used both as landmark for valve implantation and as the substrate in which to anchor the valve. Patients with NAVR tend not to have significant calcification, increasing the risk of AR or device embolization. Lastly, a significant number of patients with NAVR have an elliptical aortic annulus with a significant eccentricity index. Such asymmetric annuli can similarly set the stage for PVL despite correct positioning of the prosthesis. In addition, oversizing of the prosthesis relative to the annulus was performed using criteria derived from the AS literature. The method of valve oversizing (area- vs. perimeter-based) was reported in only 8 studies (26%). When not reported, oversizing was calculated using the data available, with preference for a perimeter-based approach, particularly with the large proportion of CoreValve devices. Additional oversizing may be required when treating patients with AR. These challenges are being addressed with newer valve designs, improved preprocedural evaluation (i.e., cross-sectional imaging) and ever-advancing operator experience.

Interestingly, self-expanding prostheses were used in the vast majority of cases. This likely reflects the greater range of sizing options and improved stability and anchoring in dilated, noncalcified aortas. However, self-expanding prostheses are known to have higher rates of AR, which may have contributed to the rates observed in the current study. To address this concern, 2nd-generation valves have various designs. For example, the JenaValve and J-Valve system rely on clip-based fixation over the native aortic valve leaflet, alleviating the dependency on aortic annular calcification. Alternatively, the ACURATE TA device utilizes a self-expandable Nitinol stent with a double polyethylene skirt that anchors against the aortic annulus. This also serves to isolate the stent from the tissue and enhances sealing to minimize PVL. The 2nd-generation valves also have the additional benefits of being recapturable and repositionable. This should improve device success, decrease the need for second valve implantation, and reduce significant PVL. Newer valves continue to be developed, including the CoreValve Evolut R,45 with recent approval of the Evolut R XL 34 mm, which should further expand the range of annuli for which percutaneous therapy can be offered – particularly relevant in the NAVR population.

Nonetheless, in this select cohort of NAVR patients with prohibitive surgical risk, TAVI proved to be a viable therapeutic option. Indeed, the reported 30-day mortality rate was similar to reported outcomes of surgical aortic valve replacement in high-risk AR patients. For example, Chaliki et al reported a 30-day operative mortality rate of 6% for patients with moderately impaired left ventricular ejection fraction (LVEF 35–50%) and 14% for patients with high surgical risk (defined as LVEF <35%).46 Moreover, when compared with TAVI performance in AS, we found comparable safety outcomes, with the exception of a higher incidence of PVL. For example, both the PARTNER- B trial2 and the US Pivotal Registry arm for patients with prohibitive surgical risk43 reported 30-day mortality rates of 5% and 8.9%, respectively. Most recently, Khatri et al performed a meta-analysis of 16,000 patients undergoing TAVI and noted a 9.1% 30-day and 21% 1-year mortality rate, nearly identical to the reported 9.6% and 21.3% found in our review.47 Thus, in select patients, TAVI for NAVR can achieve comparable outcomes to those for AS. The majority of patients with significant PVL were managed with a second TAVI-in-TAVI, with significant improvement in leakage in all but 1 case.

Procedural safety was also comparable, with a similar incidence of serious adverse events to AS. Specifically, there were no reported cases of annulus rupture and the lack of valvular calcium appears to have translated into a lower stroke rate than in either the PARTNER-B or the US Pivotal Registry (1.5% vs. 6.7% and 4.0%, respectively). Perhaps most importantly, device success (our primary outcome) was achieved in 86.3% of patients in our study, again nearly identical to the reported outcome of 84.6% in US Pivotal registry. However, when compared with patients with AS, the predominant mechanism of device failure in patients with NAVR was attributed to the need for a second TAVI valve (11.3%) and residual AR/PVL (4.6%), which has been noted to occur in as few as 1% of patients in the US Pivotal Registry. Certainly, the current data supports the safety and efficacy of TAVI for NAVR and future studies should focus on identifying strategies to minimize PVL in these patients and optimizing device success.

Study Limitations

First, TAVI for patients with isolated NAVR remains infrequent precluding us from drawing firm conclusions. Little of the analyzed data comprised consecutive patients, with the dataset derived from a large number of case reports with diverse etiologies of AR, leading our analysis to be influenced by publication bias whereby not all outcomes are equally likely to be reported. Second, the types of valves used in this study vary in design, sizing and implantation technique, as well as performance, preventing our results from being broadly applied to all types of TAVI devices. In addition, few of the studies within this work provided details of the approach to valve oversizing, specifically area- vs. perimeter-based, precluding any additional analysis to this end. Finally, the study publications spanned a period of 5 years, during which significant advancements in both the knowledge and technical skill of TAVI teams would have occurred, meaning conventional outcomes may be improved over those reported. Nonetheless, we systematically summarized the performance of TAVI in NAVR and the results provide reassurance for operators considering TAVI in these prohibitive-risk patients.

Conclusions

In conclusion, for selected patients with isolated NAVR, TAVI appears both safe and feasible with outcomes comparable to those reported for AS. A randomized study of TAVI vs. medical therapy in prohibitive-risk patients is warranted.

Acknowledgment

We thank Agnieszka Szczotka for her assistance in conducting the literature review.

Conflicts of Interest

A.L. is a consultant for Medtronic and Direct Flow Medical.

Supplementary Files

Supplementary File 1

Risk of Bias Assessment

Table S1. Excluded studies

Table S2. Case series quality: quality appraisal checklist for case series studies

Table S3. Case report quality: JBI critical appraisal checklist for case reports

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-17-0672

References
 
© 2017 THE JAPANESE CIRCULATION SOCIETY
feedback
 Top