Severe aortic stenosis (AS) is the most prevalent type of valvular heart disease in the western world,1 with particular high prevalence in older patients.2 Global demography shows a pattern of an aging population and since 1960 life expectancy at birth in Portugal has increased from 64 to 82 years. This will increase the burden of AS and the need to treat it.3,4 Surgical aortic valve (SAV) replacement is an established treatment for symptomatic severe AS.5 However, in the last decade, transcatheter aortic valve implantation (TAVI) has emerged as a minimally invasive treatment for AS, initially for inoperable and high risk patients6 but increasingly, as evidence grows, it is used in lower risk patients.7,8 Evidence is building on the durability of transcatheter aortic valves, with over 90% of patients remaining free of structural valve dysfunction between five and 10 years post-implantation.9 This may encourage increased use of this technology in younger and lower risk patients. Growing experience, solid evidence, refined pre-procedural assessment with multi-slice computed-tomography (MSCT) and new generation transcatheter devices are causing a shift toward new indications, including lower-surgical risk patients.10,11

Transfemoral (TF) access is the preferred route, since it is less invasive, and patient recovery is faster, involving shorter hospital stays.12 Nevertheless, up to one third of eligible patients may not be appropriate candidates for a TF approach because of major peripheral artery disease (PAD). Other routes have emerged as alternatives, but are associated with worse outcomes, particularly when considering the transapical (TA) route.13

Another field of increasing importance in TAVI is prosthetic valve dysfunction (PVD). Currently, patients with previously implanted biological SAV can be treated with a valve-in-valve (ViV) procedure for PVD, which has been associated with good one-year outcomes.14

Registries have reported the results of TAVI in several countries, showing consistently high implantation success rates; a recent meta-analysis reported an overall European 30-day mortality of 8%, however with substantial differences among registries.15,16 Mid- and long-term follow-up results show survival rates over 80% at one year, with persistent improvement of clinical status.17

The Portuguese National Registry of Transcatheter Aortic Valve Implantation (RNCI-VaP) is a national registry conducted with the main purpose of assessing all comers with aortic valve disease or aortic valve prosthesis dysfunction that undergo TAVI in Portugal, ensuring comprehensive coverage in Portugal.18 The main goals of the registry are to document in a prospective and consecutive fashion the characteristics of patients who undergo TAVI, ascertain the results of these treatments in Portugal, develop scientific work based on data included in the registry, establish treatment recommendations regarding TAVI procedures and collaborate with other similar international registries. A position paper statement from the Portuguese Association of Cardiovascular Intervention (APIC) has already underlined the importance of a systematic assessment and report of patients who undergo TAVI in Portugal.19 The members of this committee are all interventional cardiologists. It is divided into an executive board of three members which includes the Registry Coordinator and the scientific board. There is no specific funding for this registry which has been completely developed, maintained and sponsored by APIC.

This article is the first report of real-world clinical data from the RNCI-VaP registry. It aims to identify patient and procedure characteristics and evaluate 30-day and one-year outcomes of TAVI in Portugal.

Since its introduction in Portugal in 2007, the number of TAVI procedures has grown exponentially,18 similar to the pattern seen following the adoption of this technology in Western countries.23

Improved survival rates after TAVI, with a significant decrease in periprocedural mortality, reflect innovative device technology, but also the increased use of transfemoral TAVI, learning curves and optimized patient selection.24 The increasing availability of smaller delivery systems and operator experience have brought about an increase in the number of patients treated with a TF approach, and approximately 90% of our patients were effectively treated using TF access, a percentage slightly above the 80% reported in FRANCE TAVI, with a corresponding decline in transapical access.25 The predominant use of SEV in our registry may have, on the one hand, been influenced by the steady rise in TF TAVI, and on the other according to their availability in Portugal. As expected, balloon-expandable valves (BEV) were the main valve system used in TA TAVI. Nevertheless, it is noteworthy to mention that currently there is not enough evidence to claim the superiority of one device over the other.26 Valve selection should be tailored according to the patient's clinical and anatomical characteristics, which is standard of practice at the different Portuguese TAVI centers.

Evidence showing that ViV procedures for prosthesis dysfunction have favorable outcomes14 also contributed to the increasing number of these procedures in Portugal.

Approximately 50% of patients were considered high risk for surgery following a heart team assessment. This was not exclusively defined according to conventional surgical risk scores, in fact, in our cohort the mean EuroS-II was in the intermediate strata. Other factors, such as advanced age, associated comorbidities and frailty were determining factors, although the latter has not yet been fully reflected in the registry. Evidence shows that decision making in valvular heart disease involves a complex algorithm to define the risk-benefit ratio. Currently there is no single validated specific TAVI risk score, suggesting that the best approach should be based on the combination of the existing surgical scores together with frailty parameters and the presence of specific organ failure. The main goal should be to avoid TAVI futility.21,22

Our analysis revealed an incidence of stroke after TAVI of 4.6%, of which almost half were disabling. Although this result is below the incidence of stroke reported in the PARTNER cohorts A and B (5.5 and 6.7% respectively),27,28 and similar to the incidence of peri-operative stroke following surgical AVR in elderly patients, which ranges from 3 to 7%,29 this is still a matter of concern and an area of improvement that needs to be addressed. Embolic protection devices (EPD) are not used by default at Portuguese TAVI centers, however, so far, although published results show that EPD may help reduce the volume and size of periprocedural silent ischemic brain lesions identified on MRI, it did not reduce the incidence of new lesions and clinically significant neurological events.30,31The clinical impact of EPD needs further investigation in randomized trials prior to its widespread use.

The pacemaker implantation rate in our registry was 19% in the overall cohort, with a significant difference between TF 19,5% and TA 10.7% approach. This attributed to the differences between the predominant prosthesis type used in each of the access routes cohorts, SEV and BEV, respectively. A recent meta-analysis of different registries showed pacemakers were a requirement in 8% of Sapien TFs and 22% for CoreValve,15 in line with our data, but higher than those reported in the PARTNER trials, probably because of the higher anatomic complexity of real-world patients.

Our registry reported an overall 7.4% incidence of moderate paravalvular leak (>=2), which is similar to other European registries and explained by the operator learning curves, standardization of techniques and new generation devices with extended sealing skirts.32

Concerning hard outcomes, in the PARTNER study, mortality was 3.4 and 5% in cohorts A and B, respectively.6,33 Different registries present wide variability in mortality ranging from 5.4 to 12.4%, however, they compare different patients risk profiles, types of prosthesis and access routes.29,34–36 In our global cohort, 30-day mortality has been decreasing since the beginning of the TAVI program, from 14.1% in 2011 to approximately 4% in the last 3 years, which is in line or slightly below other reported registries.14,25 It is probably attributable to the fact that although our cohort had an advanced mean age of 81 years, the mean risk profile was intermediate. There is, in fact, a growing body of evidence that supports the extension of TAVI to increasingly lower risk patients, consistently demonstrating its noninferiority in comparison with conventional surgery.8,11 It is also important to highlight that the low mortality of our cohort evidences that most of the fatalities were unrelated to periprocedural complications, which had low prevalence. A major conclusion from these data is that our focus should be on better prediction of risks and benefits of the procedure and appropriate patient selection.

Regarding the non-TF approach, and particularly the TA, when comparing with TF route, the former was associated with a lower 30-day safety composite outcome and higher mortality in our registry, similar to what has been reported in other studies.33,37–39 The explanation is multifactorial, but clearly TA patients had increased prevalence of significant co-morbidities and a higher risk profile. A significant contribution might be derived from the increased rate of procedural complications, such as life-threatening and major bleeding, perioperative myocardial infarction and cardiac tamponade.

Based on VARC 2 criteria, the predictors for 30-day all-cause mortality were PAD, significant CAD with previous PCI, depressed left ventricular ejection fraction (LVEF) and an advanced heart failure (HF) clinical status. Buellesfeld et al. showed the patient's baseline functional state to be an independent predictor of hospital mortality40 and the Italian registry indicated that an ejection fraction under 40%, diabetes, prior valvuloplasty and procedure-related complications were associated with worse short term prognosis.29 The Ibero-American registry also underlined the importance of post-TAVI complications and logistic EuroS-II as predictors of mortality at 30 days.16 Also, in a recent meta-analysis stage>=2 AKI was a strong predictor for 30-day mortality, although interestingly many patients saw an improvement in renal function after TAVI rather than kidney injury due to renal decongestion and increased perfusion. Preprocedural pro-brain natriuretic peptide levels were also a strong predictor of 30-day and midterm mortality. This is believed to be related to advanced HF with systolic and diastolic dysfunction. Likewise, baseline clinical status, represented by prior hospitalizations and preprocedural LVEF<=30%, had a significant impact on short term mortality.41

As independent predictors of one-year all-cause mortality we identified patient characteristics before the procedure, such as pre-procedural NYHA class III/IV and chronic obstructive pulmonary disease, and procedural factors such as the non-NF access route, which may be a surrogate for significant PAD. These results are corroborated by other registries and trials, showing also that some comorbidities are related to worse long-term prognosis, such as advanced age,17 frailty,35 male gender, end-stage renal disease,34 porcelain aorta,35 low gradient aortic stenosis,42 reduced LVEF,17 NYHA class,36 mitral insufficiency grade>=2+,42 residual postintervention aortic regurgitation,6 intraprocedural conversion to surgery, peri-intervention stroke,39 or postintervention pulmonary embolism.42

In our registry there was a sub-optimal percentage of patients lost to follow-up at one-year (22%), although acceptable when compared to other observational studies.43,44,45 This can be explained by the nature of this prospective nationwide real-world registry that involves participating centers with various differences relating to the volume of procedures, follow-up approach and data registration. Also, compared to other European countries, the small number of 0.346 to 0.7 TAVI centers per million inhabitants in the registry history, can limit patient accessibility to those hospital centers because of distance, which is a known predictor of loss to follow-up.47 It is worth mentioning that an automatic registration of a patient's vital status in the RNCI-VaP could be an effective way of strengthening the quality and completeness of data in the future.

Despite the acknowledged potential bias, the findings from this landmark study should help us to select appropriate candidates for TAVI accurately, with the main goal of referring patients in whom treatment benefit outweighs the risk, thus improving clinical benefit and avoiding futility. Nevertheless, a balance needs to be achieved because even in high risk patients with several comorbidities, health- related improvements in quality of life and reduced congestive HF hospitalizations should be taken in account.

The data from the RNCI-VaP registry is self-reported and consequently subject to inclusion bias. Nevertheless, this registry is representative of the current standards of practice in Portugal, because it obtained data on 96.5% of the total number of implanted valves.

For some relevant variables we could not obtain data from all patients, which is a limitation inherent to the characteristics of a national registry that depends on the voluntary participation of the investigators, potentially causing selection bias. Events were reported by the investigators and were not independently reviewed, so we cannot rule out the possibility of under or over reporting of events. Also, the RNCI-VaP does not currently have either internal or external auditing. As such, definitive conclusions regarding cause and effect cannot be drawn from this study.

Although we acknowledge these limitations, the investigators of the RNCI-VaP consider this registry to be essential to ensure the safe and successful clinical applicability of the constantly evolving transcatheter aortic valve technology in Portugal. This paper includes the first published data on patients treated with TAVI in Portugal. It is noteworthy that the success and complication rates were similar to those reported by other European countries’ registries.32

The RNCI-VaP shows that TAVI has been implemented with great success in Portugal. In this real-world all-comers all-centers registry of patients who underwent TAVI (both conventional and ViV), the procedure was safe and effective. Characteristics related to the access route, patient comorbidities and baseline clinical status adversely affected long-term prognosis and should therefore be taken in consideration when selecting patients for the procedure.

The systematic assessment and reporting of patients who underwent TAVI in Portugal is of paramount importance, eliciting a pragmatic assessment of TAVI results in the different centers and stimulating scientific multicentric production.

CG, PCF and RCT: study concept, design and drafting of the manuscript; CG, PCF, SM, RCT: analysis and interpretation of data; PB, PCS, LP, JCS, JB, MA, VGR, BS, JB, EIO, DC, JS: critical revision of the manuscript for important intellectual content.

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

The authors have no conflicts of interest to declare.