OUP user menu

★ Editor's choice

Transcatheter aortic valve implantation: 3-year outcomes of self-expanding CoreValve prosthesis

Gian Paolo Ussia, Marco Barbanti, Anna Sonia Petronio, Giuseppe Tarantini, Federica Ettori, Antonio Colombo, Roberto Violini, Angelo Ramondo, Gennaro Santoro, Silvio Klugmann, Francesco Bedogni, Francesco Maisano, Antonio Marzocchi, Arnaldo Poli, Marco De Carlo, Massimo Napodano, Claudia Fiorina, Federico De Marco, David Antoniucci, Emanuela de Cillis, Davide Capodanno, Corrado Tamburino
DOI: http://dx.doi.org/10.1093/eurheartj/ehr491 969-976 First published online: 12 January 2012

Abstract

Aims The paucity of evidences about the long-term durability of currently available transcatheter prostheses is one of the main issues of transcatheter aortic valve implantation (TAVI). We sought to assess 3-year clinical and echocardiographic outcomes of patients undergoing TAVI with the third generation CoreValve prosthesis (Medtronic Incorporation, MN, USA).

Methods and results From the Italian CoreValve registry, 181 who underwent TAVI from June 2007 to August 2008 and eligible for 3-year follow-up were analysed. All outcomes were defined according to the Valve Academic Research Consortium. All-cause mortality at 1, 2, and 3 years was 23.6, 30.3, and 34.8%, respectively. Cardiovascular death at 1, 2, and 3 years was 11.2, 12.1, and 13.5%, respectively. The actuarial survival free from a composite of death, major stroke, myocardial infarction, and life-threatening bleeding was 69.6% at 1 year, 63.5% at 2 years, and 59.7% at 3 years. Patients with renal insufficiency had a higher mortality at 3-year follow-up (49.0 vs. 29.2%, P = 0.007); moreover, patients experiencing post-procedural major or life-threatening bleeding had a higher rate of mortality already seen at 30 days (21.6 vs. 2.8%; P < 0.001) and this result was sustained at 3-year follow-up (62.2 vs. 27.7%; P < 0.001). Mean pressure gradients decreased from 52.2 ± 18.1 mmHg (pre-TAVI) to 10.3 ± 3.1 mmHg (1-year post-TAVI) (P < 0.001); aortic valve area increased from 0.6 ± 0.2 cm2 (pre-TAVI) to 1.8 ± 0.4 cm2 (1-year post-TAVI); these results remained stable over the 3 years of follow-up. Paravalvular leak was observed in the majority of patients. There were no cases of progression to moderate or severe regurgitation. No cases of structural valve deterioration were observed.

Conclusion This multicentre study demonstrates that TAVI with the 18-Fr CoreValve ReValving System is associated with sustained clinical and functional cardiovascular benefits in high-risk patients with symptomatic aortic stenosis up to 3-year follow-up. Non-cardiac causes accounted for the majority of deaths at follow-up.

  • Aortic stenosis
  • TAVI
  • Heart valve prosthesis
  • Durability

See page 947 for the editorial comment on this article (doi:10.1093/eurheartj/ehs018)

Introduction

Transcatheter aortic valve implantation (TAVI) is an emerging, catheter-based technology that allows for implantation of a prosthetic valve without open heart surgery for the treatment of severe aortic stenosis (AS).1 The recently published Placement of Aortic Trans Catheter Valves (PARTNER) randomized controlled trial cohort B demonstrated that TAVI remarkably reduced the mortality, when compared with standard therapy, in patients not suitable candidates for surgery.2 In addition, PARTNER cohort A results showed that 1-year outcomes after TAVI compare favourably with surgical aortic valve replacement.3 While a number of large single-arm registries demonstrated encouraging short and mid-term effectiveness of TAVI,49 there is a paucity of data on benefits of this technique at longer follow-up.10,11 The lack of evidences for the long-term durability of currently available transcatheter heart valves is one of the main issue which prevents TAVI to be used in younger and lower risk patients. The aim of this multicentre study was to assess 3-year clinical and echocardiographic outcomes in patients undergoing TAVI with the third generation (18-Fr) CoreValve prosthesis (CRS) (Medtronic Incorporation, MN, USA).

Methods

Study design and patient population

Consecutive patients undergoing TAVI with the 18-Fr CRS at 12 centres across Italy and enrolled prospectively in a dedicated web-based database from June 2007 to August 2008 formed the study population. Patient eligibility criteria, registry design, features of the third generation CRS, and technical details of the procedure have been described elsewhere.7,12,13 Briefly, the principal investigators of the first four centres who began CoreValve implantation in Italy (C.T., G.T., A.S.P., F.E.) designed the study as part of their role on the steering committee, in collaboration with the sponsor, Endotech (CO, Italy). The sponsor financed the web-based database and was involved in collection, source verification, and quality control of the data, with oversight by an independent clinical events committee. All patients had severe symptomatic AS with valve area <1 cm2. Eligibility for TAVI was established at each centre based on the consensus of a local multidisciplinary team, including clinical cardiologists, cardiac surgeons, and cardiac anaesthesiologists.14 All the procedures were approved for compassionate use in patients considered at high risk for surgery, according to accepted inclusion and exclusion criteria.15 Written informed consent was obtained in all cases. Clinical follow-up was planned at 1, 3, 6, and 12 months and then yearly with visits or telephone contacts. Echocardiographic follow-up was not mandatory, but recommended at 1, 3, 6, and 12 months and then yearly.

Definitions

Device success, cardiovascular death, peri-procedural and spontaneous myocardial infarction, strokes, bleedings, combined safety and efficacy endpoints, and echocardiographic criteria post-TAVI were defined according to the Valve Academic Research Consortium (VARC).16 All major neurological events were adjudicated by the cardiologist, anaesthesiologist, and specialized radiologist with the aid of computed tomography imaging.

Statistical analysis

Continuous variables were analysed for a normal distribution with the Shapiro–Wilk test. Continuous variables following a normal distribution are presented as mean ± standard deviations and were compared using repeated measures ANOVA for within-group comparisons. Variables not following a normal distribution are expressed as median (IQR). Categorical variables are presented as counts and percentages. The cumulative incidences of clinical events at follow-up were assessed with the Kaplan–Meier method, and the log-rank test was used for comparison between patients with renal insufficiency and those experiencing bleeding and the rest of the study population. All data were processed using the Statistical Package for Social Sciences, version 15 (SPSS, Chicago, IL, USA). The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agreed to the manuscript as written.

Results

Among all patients comprised in the registry, 181 eligible for 3-year follow-up formed the study cohort. Baseline demographic, clinical and echocardiographic variables of these patients are presented in Table 1. Mean age was 80.9 ± 6.1 years and 55.8% of patients were female. All patients had severe symptomatic AS [mean aortic valve area [AVA] 0.61 ± 0.23 cm2). Overall, the population was at high surgical risk with a predicted 30-day mortality of 24.0 ± 13.5% by Logistic EuroScore17 and 11.4 ± 9.9% by STS-mortality score. The majority of patients (N = 124; 68.5%) were in NYHA functional class 3 or 4 before the procedure. Trans-femoral access was used in 172 patients (95.0%); in 9 patients (5.0%) where the trans-femoral approach was unfeasible, a trans-subclavian access was employed. Device success was obtained in 166 patients (91.7%). The prostheses' size and others procedural variables and complications are listed in the Table 2.

View this table:
Table 1

Baseline characteristics

Overall population (n = 181)
Age, years ± SD80.9 ± 6.1
Female gender, n (%)101 (55.8)
Diabetes mellitus, n (%)59 (32.6)
Coronary artery disease, n (%)96 (53.0)
Prior acute pulmonary oedema, n (%)62 (34.2)
Prior balloon valvuloplasty, n (%)34 (18.8)
Prior myocardial infarction, n (%)43 (23.8)
Prior stroke, n (%)8 (4.4)
Prior bypass graft surgery, n (%)34 (18.8)
Prior percutaneous coronary intervention, n (%)51 (28.2)
Peripheral vascular disease, n (%)27 (14.9)
Chronic obstructive pulmonary disease, n (%)34 (18.8)
Cirrhosis Child class A or B, n (%)6 (3.3)
Prior neoplasia, n (%)28 (15.5)
Renal insufficiencya, n (%)52 (28.7)
Atrial fibrillation, n (%)23 (12.7)
Prior pacemaker, n (%)16 (8.8)
Porcelain aorta, n (%)39 (21.5)
NYHA class III and IV, n (%)124 (68.5)
Logistic Euroscore, % ± SD24.0 ± 13.5
STS Score mortality, % ± SD11.4 ± 9.9
  • NYHA, New York Heart Association; STS, Society of Thoracic Surgeons.

  • aDefined as serum creatinine >1.5 mg/dL.

View this table:
Table 2

Procedural variables

Overall population (n = 181)
Procedural variables
 Procedure time, min ± SD68.6 ± 28.4
 Fluoroscopy time, min ± SD17.9 ± 8.7
 Approach
  Trans-femoral, n (%)a172 (95.0)
  Trans-subclavian, n (%)9 (5.0)
 Anaesthesia
  Local, n (%)103 (56.9)
  General, n (%)78 (43.1)
 Deviceb
  CRS 26 mm, n (%)107 (59.1)
  CRS 29 mm, n (%)74 (40.9)
 Device success, n (%)c166 (91.7)
 Post-dilatation, n (%)18 (9.9)
 Reposition with snaring, n (%)1 (0.5)
 Valve-in-valve, n (%)8 (4.4)
  Higher implantation, n (%)1 (0.5)
  Lower implantation, n (%)7 (3.9)
 Valve-on-valve, n (%)0 (0.0)
Procedural complications
 Major vascular complications, n (%)c6 (3.3)
 Percutaneous treatment, n (%)3 (1.7)
  Covered stent, n (%)2 (1.1)
  Not-covered stent, (%)1 (0.5)
  Surgical treatment, n (%)2 (1.1)
  Conservative treatment, n (%)1 (0.5)
 Procedural MI, n (%)c8 (4.4)
 RBC Units transfusions ≥4, n (%)11 (6.1)
RBC Units transfusions ≥2 and <4, n (%)40 (22.1)
  • CRS, CoreValve Revalving System; MI, myocardial infarction; RBC, red blood cells.

  • aAll cases with totally percutaneous access.

  • bRefers to the first prosthesis implanted.

  • cAccording to VARC definitions.

3-year clinical outcomes

Clinical follow-up was available in 178 patients (98.3%) at a mean of 41 ± 3 months (range 36–51 months) after TAVI. All-cause mortality rates at 1, 2, and 3 years were 23.6, 30.3, and 34.8%, respectively (Figure 1A). Cardiovascular mortality rates at 1, 2, and 3 years were 11.2, 12.1, and 13.5%, respectively (Figure 1B). As showed in the Table 3, a total of 67 (37.0%) patients died during the follow up period and the majority of deaths were not related to cardiovascular events. Cardiovascular death occurred at a median of 1.2 months (IQR 0.8–9.2) and non cardiovascular death occurred at a median of 11.6 months (IQR 2.5–25.1). Rates of VARC-defined neurologic events, acute kidney injury, myocardial infarction, bleedings and hospitalization due to cardiac reasons are listed in the Table 4. Patients experiencing an in-hospital admission due to cardiac causes at 3-years had moderate to severe mitral regurgitation in 52% or systolic pulmonary artery pressure >60 mmHg in 44% of cases.

View this table:
Table 3

Causes of mortality up to 3 years

30 days (n = 20)1 year (n = 42)3 years (n = 62)
Respiratory (pneumonia, chronic obstructive pulmonary disease exacerbation, respiratory insufficiency, acute embolism), n (%)3 (15.0)9 (17.2)13 (20.9)
Heart failure and cardiogenic shock, n (%)7 (35.0)9 (17.2)11 (17.7)
Neurological (haemorrhagic and ischaemic stroke, cerebral oedema, neurological deterioration), n (%)3 (15.0)6 (14.3)10 (16.1)
Cancer, n (%)0 (0.0)2 (4.8)5 (8.1)
Acute renal failure, n (%)1 (5.0)3 (7.1)4 (6.4)
Sepsis, n (%)0 (0.0)1 (2.4)3 (4.8)
Vascular causes, n (%)3 (15.0)3 (7.1)3 (4.8)
Gastrointestinal (GI bleedings, acute cholecystitis), n (%)1 (5.0)2 (4.8)2 (3.2)
Acute liver failure, n (%)1 (5.0)2 (4.8)2 (3.2)
Natural causes, n (%)0 (0.0)1 (2.4)2 (3.2)
Sudden death, n (%)0 (0.0)1 (2.4)1 (1.6)
Unknown, n (%)0 (0.0)2 (4.8)3 (4.8)
Others, n (%)1 (5.0)1 (2.4)3 (4.8)
View this table:
Table 4

Overall outcomes of patient population according to VARC definition

30-day outcomes
 All cause death, n (%)20 (11.2)
 Cardiovascular death, n (%)12 (6.7)
 Procedural MI, n (%)8 (4.5)
 Spontaneous MI, n (%)1 (0.6)
 Major stroke, n (%)5 (2.8)
 Life-threatening bleeding, n (%)15 (8.4)
 Major bleeding, n (%)19 (10.7)
 AKI stage 1, n (%)10 (5.6)
 AKI stage 2, n (%)11 (6.2)
 AKI stage 3, n (%)12 (6.7)
 AKI requiring renal replacement therapy, n (%)4 (2.2)
 PM implantation, n (%)22 (12.1)
 Combined safety endpoint, n (%)46 (25.8)
1-year outcomes
 All cause death, n (%)42 (23.6)
 Cardiovascular death, n (%)20 (11.2)
 Spontaneous MI, n (%)2 (1.1)
 Major stroke, n (%)6 (3.4)
 Life-threatening bleeding, n (%)17 (9.5)
 Major bleeding, n (%)19 (10.7)
 In-hospital admission, n (%)a14 (7.6)
 Re-intervention for valve dysfunction, n (%)0 (0.0)
 Combined efficacy endpoint, n (%)42 (26.6%)b
3-year outcomes
 All cause death, n (%)62 (34.8)
 Cardiovascular death, n (%)23 (12.5)
 Spontaneous MI, n (%)2 (1.1)
 Major stroke, n (%)7 (3.9)
 Life-threatening bleeding, n (%)18 (10.1)
 Major bleeding, n (%)19 (10.7)
 In-hospital admission, n (%)a24 (34.4)
 Re-intervention for valve dysfunction, n (%)0 (0.0)
 Combined efficacy endpoint, n (%)59 (37.3)b
  • Data reported as Kaplan–Meier event rate.

  • VARC, Valve Academic Research Consortium; MI, myocardial infarction; AKI, acute kidney injury; PM, pace maker.

  • aDue to cardiac reasons.

  • bPercentage of 158 patients alive after 30 days.

Figure 1

Kaplan–Meier percentage of (A) mortality rate, (B) rate of cardiovascular mortality, and (C) rate of a composite of death, major stroke, myocardial infarction, and life-threatening bleeding at 3-year follow-up.

The actuarial rate of a composite of death, major stroke, myocardial infarction and life-threatening bleeding was 30.1% at 1 year, 36.5% at 2 years, and 40.3% at 3 years (Figure 1C). Overall, the VARC-defined combined efficacy endpoint at 3-years was achieved in 59 patients among the 158 who were alive at 30-day (57.3%). Rates of major stroke and bleedings (including life-threatening and major bleedings) were 2.8 and 20.4% at 30 days and 3.9 and 22.1% at 3 years. Figure 2 shows the Kaplan-Meier curves for mortality stratified by the presence of renal insufficiency and post-procedural bleedings (including life-threatening and major bleedings). Patients with renal insufficiency did not present significant differences in term of 30-day mortality (3.8 vs. 7.8%; P = 0.332) compared with those without, whereas they had higher mortality at 3-year follow-up (51.0 vs. 29.2%, P = 0.007); on the other hand, patients experiencing post-procedural major or life-threatening bleeding had a higher rate of mortality already at 30 days (21.6 vs. 2.8%; P < 0.001) and this result was sustained at 3-year follow-up (62.2 vs. 28.4%; P < 0.001).

Figure 2

Kaplan–Meier curves of mortality at 3-year follow-up for patients with renal insufficiency (A) and patients who had post-procedural major or life-threatening bleeding (B).

NYHA class 3/4 at 1-year was virtually absent in the study population (6%). This benefit was sustained at 2- and 3-year. (Figure 3).

Figure 3

New York Heart Association (NYHA) functional status at baseline and up to 3-year follow-up. In the black box, the percentage of patients death at each timeframe.

3-year prosthesis performance

Echocardiographic follow-up was available for 89 of 119 patients (78.1%) who were alive at 3 years. Mean pressure gradients decreased from 52.2 ± 18.1 mmHg (pre-TAVI) to 10.3 ± 3.1 mmHg (1 year post-TAVI) (P < 0.001); AVA increased from 0.6 ± 0.2 cm2 (pre-TAVI) to 1.8 ± 0.4 cm2 (1-year post-TAVI); these results remained stable over the 3 years of follow-up (Figure 4). Paravalvular leak was observed in the majority of patients. Mostly, it was trace to mild in at each time point (48.1, 50.3, and 47.2% at 1, 2, and 3 years, respectively); in detail, moderate paravalvular leak was observed in 17.8, 16.5, and 10.1% of patients at 1, 2, and 3 years, respectively (Table 5). All patients with mild paravalvular leak were either unchanged or improved over time. There were no cases of progression to moderate or severe regurgitation. No cases of structural valve deterioration were observed.

View this table:
Table 5

Baseline, postprocedural, and up to 3-year echo parameters

VariablesBaseline (n = 181)Post-procedure (n = 178)1 year (n = 129)3 years (n = 89)
Mean gradient, mean ± SD, mmHg52.2 ± 18.110.3 ± 3.910.3 ± 3.110.3 ± 4.7
Mean gradient >20 mmHg, n (%)181 (100)0 (0.0)1 (0.8)1 (1.1)
LV-EF mean ± SD, %51.3 ± 13.152.4 ± 11.554.1 ± 9.951.8 ± 13.9
Aortic regurgitationa, n (%)14 (7.7)NANANA
Periprosthetic leak, n (%)
 Mild PPL, n (%)NA94 (52.8)62 (48.1)42 (47.2)
 Moderate PPL, n (%)NA27 (15.2)23 (17.8)9 (10.1)
 Severe PPL, n (%)NA0 (0.0)0 (0.0)0 (0.0)
  • LV-EF, left ventricle ejection fraction, PPL, peri-prosthetic leak; NA, not available.

  • aExceeding grade 2/4+.

Figure 4

Time trends in transaortic mean gradient and aortic valve area.

Discussion

The introduction of TAVI for treatment of severe AS has expanded the therapeutic possibilities for successfully managing severe AS. Although large registries and randomized trials have proved the effectiveness of this procedure, TAVI is still confined to the treatment of elderly and high-risk patients, because of the newness of the technique, which prevents to make reliable estimation on long-term durability of both available CE mark prostheses. The present multicentre analysis from the Italian registry describes the outcomes of a cohort of TAVI patients with the longest follow-up using the self-expanding CoreValve Revalving System device published so far. In addition, this is the first study in which long-term outcomes are defined according to the standardized definitions of the VARC.

The 23.6% 12-month mortality in this study compares favourably to other balloon-expandable and self-expanding TAVI clinical reports, including PARTNER trial cohorts A and B and recent national registries, which showed mortality at the same timeframe ranging from 15 to 30.29 However, the central finding of the present analysis is that 3-year survival after TAVI with the CRS prosthesis amounted to 65%. These data are not surprising, given the baseline risk profile of the subjects undergoing the procedure during the early TAVI experience which was extremely high. Therefore, patients died during follow-up either because of their comorbidities or secondary to conditions associated with advanced age, as proved by the 13% of cardiovascular death rate at 3 years, which mostly occurred during the first month after TAVI, due to procedural complications. This poses an obvious challenge for longer-term real effectiveness of TAVI in such population and raises questions about whether this procedure could improve long-term results in lower-risk patients, as actually suggested by recent reports.9,18 According to our findings, since three out of five patients died due to non-cardiovascular reasons at a median of 11 months, almost half of those patients should be denied the procedure in agreement of 2008 European position statement on TAVI, which pointed out that this procedure should not be performed in patients whose life expectancy is <1 year.15 Therefore, in the future, it will be crucial to identify patient populations that can derive most advantage from the application of this technology in the perspective of a cost-benefit ratio.

Buellesfeld et al.11 reported data from the CoreValve ‘safety and efficacy’ study. Survival at 1 and 2 years was 71.9 and 59.7%, respectively; in a further analysis, the authors found that patients classified at high risk before the intervention had twice the risk of major adverse events up to 2 years compared with patients in the moderate-risk group. Although there was no difference in cardiac deaths between the moderate-risk and the high-risk operable groups, there was a significant difference in all-cause mortality due to more non-cardiac events in the high-risk population. Those findings compare favourably with those of our study as stated above about cardiovascular mortality. Compared with the report by Buellesfeld et al.,11 it has to be highlighted the lower incidence of stroke reported in our population (13.5% at 2 years vs. 3.9% at 3 years in the present series). Definitely, the different definitions of stroke between the two studies might strongly affect such comparison. On the other hand, we described a stroke rate which compares favourably with those reported by Gurvitch et al.19 using the VARC definitions.

Recently, Gurvitch et al.10 performed an analysis on long-term (3 years) outcomes and durability of Edwards SAPIEN device; given this aim, patients with unsuccessful procedures or mortality within 30 days were excluded from the analysis because these cases typically reflected procedural complications, rather than issues with the transcatheter valve per se. They showed that TAVI may be accomplished with 81, 74, and 61%, survival at 1, 2, and 3 years, respectively, which is approximately in line with our results. Again, comparisons with the present analysis regarding the incidence of major cardiovascular events (stroke, myocardial infarction, and bleedings) are limited by the heterogeneity of the definitions. Finally, Moat et al.20 reported 2-year data of the UK-TAVI Registry, which collected data of both self-expanding and balloon-expandable prostheses. Two-year survival was 74%, and the presence of moderate/severe aortic regurgitation, and chronic obstructive pulmonary disease were found the only independent predictors of mortality in a multivariate model.

With regards to late morbidity, we found that a percentage slightly higher of 50% was free from serious cardiovascular events, such as death, myocardial infarction, major stroke, and life-threatening bleeding. As shown in Table 4, major adverse cardiovascular and cerebrovascular events were primarily contributed by mortality, followed by bleedings. Moreover, almost 14% of patients had hospitalization for cardiac decompensation at follow-up, and these data are consistent with those reported in PARTNER cohort A trial up to 1 year after TAVI.3 Interestingly, most of them had moderate to severe mitral regurgitation or pulmonary hypertension (sPAP > 60 mmHg) before the procedure.

Consistently with previous studies,4,7,9,19 renal insufficiency was found as important factor determining overall and late (from 30 days up to 3 years) mortality rate; in detail, there was a tendency towards a higher 30-day mortality in patients with renal insufficiency (3.8 vs. 7.8%), even if it did not reach statistical significance, probably as the reflection of a power issue. Multiple factors may account for the mechanism linking chronic kidney disease and late mortality in TAVI patients.21,22 In addition, overall survival was strongly impacted by post-procedural stroke and bleedings. Contrary to renal insufficiency, these complications affected survival already at 30 days and maintained their effect up to 3 years. Both stroke and bleedings are well-established predictors of poor outcome at follow-up.23,24 In this analysis, the 30-day incidence of stroke and bleeding of 2.8 and 20.4% and the Kaplan–Meier freedom from stroke and bleeding at 3 years of 96.1 and 77.9% demonstrate that those events mostly occurred within 30 days from TAVI and therefore they were very likely procedure-related. In accordance with results reported by Gurvitch et al.,19 we found that an important part of patients with major vascular complications required red blood cell transfusions; on the other hand, we noted that patients undergoing blood transfusions not had a clear source of bleeding; in such patients, pre-existent anaemia and excessive haemodilution due to fluid administration might be the reason for blood transfusions. Concerning neurologic events, our findings are consistent with previous European registries,7,8,9,11 but disagree with those reported in PARTNER cohort A, in which almost the 40% of neurologic events occurred from 30 days up to 1 year.3 Three possible explanations could be hypothesized: first, the higher percentage of pre-existing cerebral vascular disease (almost 28%) in the PARTNER population could play a role in the higher incidence of stroke at follow-up; second, in the PARTNER trial 24-Fr devices were used and this technical issue might explain this difference; third, in the randomized trial, the diagnosis of neurological event was made by a specialized neurologist; in this registry, such as in others European registries, cardiologists and anaesthesiologists assessed the neurological function after TAVI, thus possibly leading to an underestimation of all not clinically relevant neurological events. However, our findings underline the importance of further improving the implantation technique with the development of supra-aortic trunks protection devices25 and with the reduction of delivery catheter profiles in order to reduce the incidence of procedure-related embolic cerebral and bleeding events.

Study limitations

The present study share several important limitations with other observational registries. First, results obtained in a non-randomized fashion were reported, with the lack of comparative arms of patients with severe AS treated by surgery or medical therapy alone. Secondly, the relatively small sample size, even if these data refer to the largest TAVI-CoreValve population with 3-year follow-up published so far. Thirdly, since our data were entered prospectively into a web-based database before the introduction of VARC definitions, re-evaluation of clinical source document relating to clinical events according to the VARC criteria might cause issues with respect to events adjudication; in particular, it may result in the underestimation of the true frequency of some clinically adjudicated events; for that reason, the Steering Committee decided not to report the incidence of minor neurological and bleeding events, which did not undergo independent monitoring or objective assessment. Fourthly, the neurologic events were not adjudicated by a neurologist, probably resulting in the underestimation of the true frequency of such events. Fifthly, echo parameters were not monitored by a central Core-laboratory. Finally, follow-up on echo data and NYHA functional class was performed in a ‘survival cohort’, with death possibly exerting a competing risk that may have biased our results. Caution therefore must be applied when interpreting these data.

Conclusions

This multicentre study demonstrates that TAVI with the 18-Fr CoreValve ReValving System is associated with sustained clinical and functional cardiovascular benefits in high-risk patients with symptomatic AS up to 3-year follow-up. Non-cardiac causes accounted for the majority of deaths at follow-up.

Funding

This study was supported by Endotech (Italy).

Conflict of interest: G.P.U., A.S.P., F.E., A.R., G.S., and F.B. are proctors physician for Medtronic Incorporation; all other authors have no relevant conflicts of interests to declare.

Acknowledgements

The authors wish to thank Cristina Giannini MD, Alessandro Santo Bortone MD, Cristina Ciuca MD, Nazario Carrabba MD, Wanda Deste MD, Patrizia Aruta MD, Nedy Brambilla MD, Diego Maffeo MD, and Carla Boschetti MD for their indispensable role in clinical and echocardiographic data collection.

References

View Abstract