Author + information
- Received May 17, 2019
- Revision received September 12, 2019
- Accepted September 15, 2019
- Published online November 25, 2019.
- Sylvain Beurtheret, MDa,∗ (, )@Beurtheretsylv1,
- Nicole Karam, MD, PhDb,c,d@nickaram,
- Noemie Resseguier, MDe,
- Remi Houel, MDa,
- Thomas Modine, MD, PhDf,
- Thierry Folliguet, MD, PhDg,
- Chekrallah Chamandi, MDb,c,d,
- Olivier Com, MDh,
- Richard Gelisse, MDh,
- Jacques Bille, MDh,
- Patrick Joly, MDh,
- Nicolas Barra, MDh,
- Alain Tavildari, MDh,
- Philippe Commeau, MDi,
- Sebastien Armero, MDj,
- Mathieu Pankert, MDk,
- Michel Pansieri, MDk,
- Sabrina Siamea,
- René Koning, MDl,
- Marc Laskar, MD, PhDm,
- Yvan Le Dolley, MDa,
- Arnaud Maudiere, MDa,
- Bertrand Villette, MDa,
- Patrick Khanoyan, MDh,
- Julien Seitz, MDh,
- Didier Blanchard, MDb,c,d,
- Christian Spaulding, MD, PhDb,c,d,
- Thierry Lefevre, MDn,
- Eric Van Belle, MD, PhDo,
- Martine Gilard, MD, PhDp,
- Helene Eltchaninoff, MD, PhDq,
- Bernard Iung, MD, PhDr,
- Jean Philippe Verhoye, MD, PhDs,
- Ramzi Abi-Akar, MDt,
- Paul Achouh, MD, PhDt,
- Thomas Cuisset, MD, PhDu,
- Pascal Leprince, MD, PhDv,
- Eloi Marijon, MD, PhDb,c,d@EloiMarijon,
- Hervé Le Breton, MD, PhDw and
- Antoine Lafont, MD, PhDb,c,d
- aCardiac Surgery Department, Saint Joseph Hospital, Marseille, France
- bParis Cardiovascular Research Center, INSERM Unit 970, Paris, France
- cUniversité Paris Descartes, Sorbonne Paris Cité, Paris, France
- dAssistance Publique-Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Cardiology Department, Paris, France
- eDepartment of Biostatistics and Public Health, La Timone Hospital, Marseille, France
- fCardiac Surgery Department, Cardiologic University Hospital, Lille, France
- gDepartment of Cardiothoracic Surgery and Transplantation, University of Lorraine, Centre Hospitalier Universitaire Brabois, Vandoeuvre les Nancy, France
- hCardiology Department, Saint Joseph Hospital, Marseille, France
- iDepartment of Cardiology, Clinique des Fleurs Ollioules, Ollioules, France
- jDepartment of Cardiology, Hôpital Européen, Marseille, France
- kDepartment of Cardiology, Centre Hospitalier Henri Duffaut, Avignon, France
- lCardiology Service, Saint Hilaire Clinic, Rouen, France
- mDepartment of Cardiac Surgery, Centre Hospitalier Dupuytren, Limoges, France
- nParis South Cardio-vascular Institute, Jacques-Cartier Private Hospital, Massy, France
- oDepartment of Cardiology, University of Lille 2, Regional University Hospital Centre of Lille, National Institute of Health and Medical Research U1011, University Hospital Federation Integra, Lille, France
- pDepartment of Cardiology, La Cavale Blanche University Hospital Centre, Optimization of Physiological Regulations, Science and Technical Training And Research Unit, University of Western Brittany, Brest, France
- qCardiology Service, Rouen–Charles-Nicolle University Hospital Centre, National Institute of Health and Medical Research U644, Rouen, France
- rDepartment of Cardiology, University Hospital Department Fire and Paris-Diderot University, Public Assistance Hospitals of Paris, Bichat Hospital, Paris, France
- sThoracic and Cardiovascular Surgery Service, Pontchaillou University Hospital Centre, University of Rennes 1, Signal and Image Treatment Laboratory (LTSI), National Institute of Health and Medical Research U1099, Rennes, France
- tCardiac Surgery Department, Assistance Publique-Hôpitaux de Paris, Hôpital Européen Georges Pompidou, Cardiology Department, Paris, France
- uCardiology Department, La Timone Hospital, Marseille, France
- vCardiac Surgery Department, Sorbonne–Pierre-et-Marie-Curie University, Assistance Publique-Hôpitaux de Paris, Groupe Hospitalier de la Pitié Salpêtrière, Paris, France
- wCardiology and Vascular Diseases Service, Pontchaillou University Hospital Centre, Centre for Clinical Investigation 804, University of Rennes 1, Signal and Image Treatment Laboratory, National Institute of Health and Medical Research U1099, Rennes, France
- ↵∗Address for correspondence:
Dr. Sylvain Beurtheret, Saint Joseph Hospital, Cardiac Surgery Department, 26 bd de Louvain, 13008 Marseille, France.
Background Femoral access is the gold standard for transcatheter aortic valve replacement (TAVR). Guidelines recommend reconsidering surgery when this access is not feasible. However, alternative peripheral accesses exist, although they have not been accurately compared with femoral access.
Objectives This study compared nonfemoral peripheral (n-FP) TAVR with femoral TAVR.
Methods Using the data from the national prospective French registry (FRANCE TAVI [French Transcatheter Aortic Valve Implantation]), this study compared the characteristics and outcomes of TAVR procedures according to whether they were performed through a femoral or a n-FP access, using a pre-specified propensity score−based matching between groups. Subanalysis during 2 study periods (2013 to 2015 and 2016 to 2017) and among low/intermediate-low and intermediate-high/high volume centers were performed.
Results Among 21,611 patients, 19,995 (92.5%) underwent femoral TAVR and 1,616 (7.5%) underwent n-FP TAVR (transcarotid, n = 914 or trans-subclavian, n = 702). Patients in the n-FP access group had more severe disease (mean logistic EuroSCORE 19.95 vs. 16.95; p < 0.001), with a higher rate of peripheral vascular disease, known coronary artery disease, chronic pulmonary disease, and renal failure. After matching, there was no difference in the rate of post-procedural death and complications according to access site, except for a 2-fold lower rate of major vascular complications (odds ratio: 0.45; 95% confidence interval: 0.21 to 0.93; p = 0.032) and unplanned vascular repairs (odds ratio: 0.41; 95% confidence interval: 0.29 to 0.59; p < 0.001) in those who underwent n-FP access. The comparison of outcomes provided similar results during the second study period and in intermediate-high/high volume centers.
Conclusions n-FP TAVR is associated with similar outcomes compared with femoral peripheral TAVR, except for a 2-fold lower rate of major vascular complications and unplanned vascular repairs. n-FP TAVR may be favored over surgery in patients who are deemed ineligible for femoral TAVR and may be a safe alternative when femoral access risk is considered too high.
Since its introduction in 2002, transcatheter aortic valve replacement (TAVR) has expanded rapidly as an alternative to surgical aortic valve replacement in patients at high and intermediate procedural risk (1–4).
Femoral peripheral (FP) access is the most studied and widely used access for TAVR procedures; it allows exclusive percutaneous intervention (5). However, despite the improvement in device profiles and procedural techniques, FP access cannot be performed in approximately 10% to 15% of patients due to iliofemoral arteriopathy, tortuosity, severe calcifications, aortic aneurysm, mural thrombus, or previous vascular surgery (6). Current guidelines recommend that surgery be reconsidered in patients ineligible for FP access, mainly based on studies that assessed the safety of transapical access (1,7). However, alternative nonfemoral peripheral (n-FP) accesses were recently developed (8–11). No randomized trial has compared the outcome of TAVR according to the access site, and observational study−derived comparisons have been limited by the difference in patient characteristics between the groups, with patients with more severe disease undergoing n-FP TAVR. Although propensity-matched comparison of transthoracic and FP TAVR has shown higher rates of adverse periprocedural events and death with central or transthoracic access (e.g., transapical or transaortic) (12,13), no similar comparisons have been made between n-FP (transcarotid or trans-subclavian) and FP TAVR.
Using data from the national prospective French registry, the FRANCE TAVI (French Transcatheter Aortic Valve Implantation) registry, we compared 30-day outcomes of TAVR procedures according to whether they were performed through FP or n-FP access. Outcomes were compared after propensity score−based matching of patients with FP and n-FP interventions, with an assessment of temporal evolution in practices and outcomes during 2 critical periods (i.e., 2013 to 2015 and 2016 to 2017) and among low/intermediate-low and intermediate-high/high volume centers.
FRANCE TAVI registry
The FRANCE TAVI registry was launched in January 2013, in continuity with the France 2 Registry, on the initiative of the Groupe Atherome coronaire et Cardiologie Interventionnelle, the Interventional Cardiology working group of the French Society of Cardiology, with the participation of the French Society of Thoracic and Cardiovascular Surgery. It was approved by the Institutional Review Board of the French Ministry of Higher Education and Research and by the National Commission for Data Protection and Liberties, whose principles are in line with the General Data Protection Regulation. FRANCE TAVI is registered with ClinicalTrials.gov (Registry of Aortic Valve Bioprostheses Established by Catheter; NCT01777828).
Designed as an all-comers registry, the FRANCE TAVI registry prospectively includes all patients undergoing TAVR procedures for severe aortic valve stenosis in 50 active TAVR centers in France. The decision to perform TAVR and the choices of access site and prosthesis type are made by a multidisciplinary heart team in each participating center. Procedures and post-procedural management are performed in accordance with the routine protocol of each site. A 30-day follow-up is recommended in the case report form and is performed either on site or by telephone contact with the patient and the patient’s physician according to the participating site preference. Patients included in the registry provide written informed consent for the procedure and for anonymous processing of their data.
The FRANCE TAVI dataset is filled through a dedicated web-based interface from the French Society of Cardiology. Collected data include baseline, procedural, and outcome characteristics. The database is managed by the French Society of Cardiology, which implements regular data quality checks, including range checks and assessments of internal consistency. In cases of missing, extreme, or inconsistent values, centers are contacted and asked to verify and modify the records as appropriate.
Outcomes are site-reported and standard definitions are used to enter the data. Major complications, including valve migration or embolization, major vascular complications, and bleeding were defined according to Valve Academic Research Consortium 2 criteria (14,15). Accordingly, major bleeding was defined as overt bleeding either associated with a decrease in hemoglobin level of at least 3.0 g/dl or requiring transfusion of at least 2 U of whole blood or red blood cells, or bleeding that caused hospitalization or permanent injury, or required surgery. Unplanned vascular repair was defined as any vascular injury that led to unplanned surgery or stent implantation. Renal failure was defined as an at least 1.5 increase in serum creatinine level compared with baseline or an increase of >0.3 mg/dl (26.4 mmol/l).
Based on criteria specified by the French Ministry of Health, patients included in the registry were adults with severe aortic valve stenosis. Severe aortic valve stenosis was diagnosed as defined by international guidelines. All patients enrolled in the FRANCE TAVI registry between January 2, 2013 and December 31, 2017 were included in the study. Patients with missing data on valve type, access site, or propensity score variables (Online Table 1) were excluded from the analysis. The study flowchart is presented in Figure 1.
This was a multicentric observational study in which each center used its own technique. FP TAVR was percutaneous in most cases. It was a surgical approach in all cases of n-FP TAVR. The surgical cutdown of the trans-subclavian access (including transaxillary or distal subclavian) was performed through an infraclavicular incision respecting the brachial plexus. In case of a carotid approach, a 3- to 4-cm long, low cervical incision was performed to expose the sternocleidomastoid muscle. The jugular vein, the Vagus nerve, and the respiratory tract were identified. Theoretically, the left access for n-FP TAVR was often preferred over the right access because it provided superior coaxial alignment with the ascending aorta and optimal positioning of the prosthesis.
Conscious sedation with local anesthesia or general anesthesia was possible with all pathways.
The primary endpoint for the pre-specified propensity score−matching based comparison was procedural mortality (either in-hospital mortality or 30-day mortality). Secondary endpoints included all other in-hospital complications (Online Table 2). Of note, 30-day follow-up was available in all patients.
This report was prepared in compliance with the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) checklist for observational studies (16).
Baseline clinical characteristics were first described and compared between the 2 groups of interest (FP TAVR vs. n-FP TAVR). Quantitative variables were presented as mean ± SD (assumption of normal distribution was assessed graphically using histograms and QQ plots) and compared using Student’s t-test. Categorical variables were presented as numbers (percentages) and compared using the chi-square test when appropriate (Fisher exact test otherwise).
Among patients with peripheral TAVR (n = 25,666), baseline characteristics differences were considered to be small because standardized differences were <0.2 (median standardized difference was 0.05) between those with a complete dataset (n = 21,611) versus those with ≥1 missing data (n = 4,055), which represented 14.5% of the overall population. We performed an unadjusted comparison between n-FP and FP approaches among patients with an incomplete record to account for missing data.
Propensity score−based matching was developed using a logistic regression model that included 18 pre-procedural variables known to be related to outcomes and/or to the access site (regardless of their statistical significance, using a non-parsimonious approach) and the center volume to abolish a possible “center effect.” The center volume was divided into 4 groups using quartiles: low volume (≤67 procedures per year); low-intermediate volume (between 68 and 104 procedures per year); intermediate-high volume (between 105 and 155 procedures per year); and high volume (≥156 procedures per year). This model allowed calculation of the probability (propensity score) of n-FP TAVR access for each patient. Using the propensity score, n-FP TAVR cases were matched to FP TAVR cases. A nearest-neighbor 1:1 matching algorithm on the basis of the propensity score was applied, with a caliper width of 0.2 SD of the logit of the propensity score. Standardized differences before and after matching were estimated (with their 95% confidence interval [CI]) to assess the quality of the propensity score matching procedure. A mirrored histogram of distribution of propensity scores for n-FP TAVR (bars below the zero line) versus FP TAVR (bars above the zero line) is presented in Online Figure 1.
Logistic regression models were used to estimate the odds ratio (OR) of each endpoint for n-FP TAVR access and adjusted for the type of prosthesis implanted and the study period. The implanted prosthesis type and the study period, which were post-baseline covariates, were therefore not included in the propensity score model and were taken into account as possible confounding factors in the multivariate regression model.
To further account for the evolution in catheter size, device performance, and practices during the study period, a first subanalysis was performed with a comparison of FP to n-FP TAVR in each study period separately using the same methodology. A second subanalysis was performed with a comparison of FP TAVR to n-FP TAVR in low/intermediate-low volume centers and intermediate-high/high volume centers separately using the same methodology. To ensure that the overall outcomes were not driven by one compared with the other n-FP approach, we performed an unadjusted comparison between transcarotid and trans-subclavian approaches.
All tests were 2-sided, and p values <0.05 were considered to be statistically significant. Statistical analyses were performed using R software version 3.4.1 (R Foundation, Vienna, Austria). The R package MatchIt was used for the propensity score−based matching.
During the study period, 27,997 patients were included in the FRANCE TAVI registry. Of those, 352 had missing data regarding the access site, 1,979 underwent a central access (transapical or transaortic), and 4,055 had at least 1 missing data among propensity score variables. Baselines characteristics and impact of access type on the outcome of the unmatched excluded population are presented in Online Tables 3 and 4. Therefore, 21,611 patients underwent peripheral vascular access and were included in the study.
Characteristics of the population before matching
FP access was used in 19,995 patients (92.5%), whereas n-FP access was used in 1,616 cases (7.5%). Baseline characteristics of the nonmatched population are presented in Online Table 5.
Patients in the n-FP access group were younger than those in the FP access group (81.83 years vs. 83.34 years; p < 0.001) and were more frequently men (64.60% vs. 47.13%; p < 0.001). Patients in the n-FP access group had more severe disease. Their mean logistic EuroSCORE was higher (19.95 vs. 16.99; p < 0.001), their ejection fractions were lower (54.45% vs. 56.48%; p < 0.001), with higher rates of known coronary artery disease (63.37% vs. 46.41%; p < 0.001), previous cardiac surgery (21.78% vs. 13.06%; p < 0.001), diabetes mellitus (28.71% vs. 25.64%; p = 0.007), renal failure (51.92% vs. 47.37%; p < 0.001), peripheral vascular disease (50.25% vs. 18.69%; p < 0.001), and chronic pulmonary disease (25.43% vs. 18.04%; p < 0.001).
Forest plots of patients’ factors associated with n-FP TAVR versus factors associated with FP TAVR are presented in Figure 2.
Characteristics of the matched population
Baseline characteristics of the matched population according to access type (n = 1,613 in each group) are shown in Table 1. In the n-FP access group, mean age was 81.85 years versus 82.08 years in the FP group (p = 0.382), with 64.54% men versus 63.30% men (p = 0.453). A nonelective procedure was performed in 8.31% of cases versus 8.68% of cases (p = 0.705). Stroke history was present in 12.09% of cases versus 13.64% of cases (p = 0.189). Overall, patients' characteristics were similar for all variables included in the propensity score. Balloon expandable prostheses were the most frequent prostheses used in the FP access group (64.41%), whereas self-expandable prostheses were the most frequently used in the n-FP access group (66.58%). Among the n-FP access group, 911 (56.5%) patients underwent transcarotid access and 702 (43.5%) patients underwent trans-subclavian access.
Impact of access site on the outcome
Impact of access type on outcomes in the unmatched population is presented in Online Table 6. There were more strokes (OR: 1.65; 95% CI: 1.22 to 2.20; p = 0.002), acute renal failures (OR: 1.39; 95% CI: 1.05 to 1.81; p = 0.024), and major bleedings (OR: 1.22; 95% CI: 1.01 to 1.47; p = 0.040), but less annulus ruptures (OR: 0.12; 95% CI: 0.00 to 0.84; p = 0.027), major vascular complications (OR: 0.55; 95% CI: 0.28 to 0.95; p = 0.031), and unplanned vascular repairs (OR: 0.45; 95% CI: 0.33 to 0.59; p < 0.001) in the n-FP group.
Among the 1,613 matched n-FP cases, mean post-procedural length of stay was significantly higher in the n-FP group (8.86 days vs. 7.98 days; p = 0.006). The procedural mortality rate was 3.97% versus 2.91% in the FP group (p = 0.211); the ST-segment elevation myocardial infarction rate was 0.25% versus 0.19% (p = 0.774), the stroke rate was 3.35% versus 2.17% (p = 0.156), the major vascular complications rate was 0.68% versus 1.36% (p = 0.032), and unplanned vascular repairs occurred in 3.10% of cases versus 6.70% of cases (p < 0.001). Compared with FP access, n-FP access was not associated with increased procedural mortality (OR: 1.29; 95% CI: 0.87 to 1.94; p = 0.211) or associated with an increased risk of stroke (OR: 1.38; 95% CI: 0.88 to 2.19; p = 0.156). As shown in Table 2, there was no difference in the rate of any complications according to the access site in the matched population, except for a 2-fold lower rate of major vascular complications (OR: 0.45; 95% CI: 0.21 to 0.93; p = 0.032) and unplanned vascular repairs (OR: 0.41; 95% CI: 0.29 to 0.59; p < 0.001) in the n-FP access group. In multivariate logistic regression, FP access was an independent predictor of major vascular complications (Online Table 7).
In the n-FP group, unmatched comparison between the transcarotid and trans-subclavian accesses (Online Tables 8 and 9) showed no difference in the stroke rate (3.62% and 2.99%, respectively; p = 0.485). There were more renal failures (OR: 1.78; 95% CI: 1.03 to 3.16; p = 0039) and major bleedings (OR: 1.82; 95% CI: 1.25 to 2.68; p = 0.002) but less hemorrhagic shocks (OR: 0.26; 95% CI: 0.06 to 0.91; p = 0.035) and major vascular complications (OR: 0.21; 0.04 to 0.80; p = 0.021).
Over the 5-year study period, the distribution of TAVR accesses significantly evolved (p < 0.001). The rate of FP TAVR increased from 79.95% in 2013 to 2015 to 89.12% in 2016 to 2017, with a massive decrease in central TAVR (i.e., transapical and transaortic) from 11.99% to 3.76%, whereas the rate of n-FP TAVR remained stable from 7.66% to 6.62% (Figure 3). Distribution of prostheses types across the 2 study periods are shown in Online Table 10.
In the total population (n = 21,611), mean post-procedural length of stay decreased in both groups (9.99 to 7.96 days in the n-FP group; p < 0.001; and 8.71 to 6.80 days in the FP group; p < 0.001). The procedural mortality rate decreased from 3.76% to 2.36% (p < 0.001), and the major vascular complications rate decreased significantly in the total population from 1.44% to 1.02% (p = 0.005). There was a nonsignificant decrease in stroke rate in the total population from 2.06% to 1.81% (p = 0.173).
When considering the 2016 to 2017 period separately, there was no difference in the rate of any complications according to the access site in the matched population, except for a 4-fold lower rate of major vascular complications (OR: 0.26; 95% CI: 0.07 to 0.78; p = 0.015) and a 2-fold lower rate of unplanned vascular repairs (OR: 0.47; 95% CI: 0.28 to 0.77; p = 0.002) in the n-FP access group. The complete per-period subanalysis is shown in Online Table 11.
Per-center volume subanalysis
In high-intermediate/high volume centers, there was no difference in the rate of complications according to the access site in the matched population, except for a 3-fold lower rate of major vascular complications (OR: 0.35; 0.13 to 0.84; p = 0.018) and a 2-fold lower rate of unplanned vascular repairs (OR: 0.44; 95% CI: 0.29 to 0.66; p < 0.001) in the n-FP access group. The per-center volume subanalysis is shown in Online Table 12.
In this large multicentric study that included all peripheral vascular TAVRs performed in France between 2013 and 2017, after a pre-specified propensity-based matching, the complications rate was low and similar between n-FP and FP TAVR, except for a 2-fold lower rate of major vascular complications or unplanned vascular repairs in the n-FP TAVR group (Central Illustration). The results were consistent over the 2 study periods, despite a reduction in procedural mortality, stroke, major vascular complications, and unplanned vascular repairs rates. The comparisons of outcomes provided similar results in intermediate-high/high volume centers.
TAVR has widely developed during the last several years, shifting from a technique that was solely used in patients who were ineligible for surgery, to a procedure that can now be considered in a large subset of patients, including those at lower risk. More recent studies have shown similar or even lower rates of death, stroke, or rehospitalization, compared with surgery, including among patients at low surgical risk (17,18). FP access is the current gold standard for TAVR procedures and is the first-line access site considered when TAVR is envisaged. Current guidelines state the feasibility of FP access is one of the key features that need to be assessed before choosing between TAVR and surgical valve replacement (1,7). However, despite great improvement in TAVR techniques and device profiles, approximately 10% to 15% of patients are still denied FP access due to unfavorable anatomy (6). In cases of intermediate-risk patients, European guidelines recommend that a surgical option be reconsidered (1). However, n-FP accesses have emerged as alternatives to FP, although no dedicated devices have been made for n-FP access (8,9,18). Those n-FP accesses have not been accurately compared with surgery or FP TAVR.
From an organizational point of view, n-FP TAVR is more invasive and more demanding than FP TAVR, in which procedural duration has decreased with time and general anesthesia can now be avoided in experienced centers (19). This led to the development of minimalistic TAVR with a shorter length of hospital stay and a simplification of procedure organization (20). Regarding safety and outcome, comparison of access types in observational studies has been limited by the great variability among patients who undergo different accesses, with patients with more severe disease undergoing n-FP TAVR. In our study, we performed propensity-based matching to allow an accurate comparison between groups. We did not observe any difference in procedural mortality or complication rates between FP and n-FP access, but n-FP access was associated with a lower rate of major vascular complications and unplanned vascular repairs compared with FP access. This confirmed the findings of previous studies (21,22).
The absence of increased risk of disabling stroke in n-FP TAVR was another major finding in our study. Despite all the advances in TAVR techniques, stroke remained the most feared complication, with evidence of silent and apparent microembolism in 50% to 94% of patients within the first month after TAVR (22,23). Cerebrovascular events were mainly attributed to the dislodgment of calcified debris and/or aortic valve tissue during the TAVR procedure (24). This hypothesis was further supported by the reduction of cerebral lesion volumes in diffusion-weighted magnetic resonance imaging in patients who underwent TAVR with embolic protection devices (25). In the case of n-FP access (mainly transcarotid), a specific mechanism for cerebrovascular events could be imagined, related to local complications and to the transient reduction in cerebral blood flow during the procedure. This theoretical increase in stroke rate was not observed in our study, which further supported a wider use of this access when FP access was deemed not feasible during the pre-operative workup.
Improvements in device performance and operator experience led to a decrease in length of stay after the procedure and a reduction in the rates of death, stroke, major vascular complications, and unplanned vascular repairs. However, comparison between FP and n-FP access provided similar results in both study periods, regardless of volume center size. In particular, n-FP access remained associated with a lower rate of unplanned vascular repairs and a lower rate of major vascular complications, which highlighted the necessity of considering n-FP access in complex FP cases. Further studies and specific scores are needed to identify the patients at higher risk for major vascular complications or unplanned vascular repairs who would benefit from this alternative access.
Despite the large number of patients included and the involvement of several centers in the study, several limitations have to be acknowledged. First, although propensity matching was performed to reduce indication bias, which allowed a more reliable comparison of patients according to the access site, the results had to be analyzed with caution because of the nonrandomized nature of the study. Some differences persisted between the groups. In particular, the most frequent prosthesis type differed according to the access site because we chose to match on only the baseline characteristics of the patients. However, the prosthesis type was adjusted in the multivariate regression model and was not likely to influence the results. Second, for the sake of matching, we excluded patients with incomplete data. However, unadjusted comparison between n-FP and FP approaches among excluded patients with incomplete data showed similar results, indicating that this did not introduce any bias. Third, only symptomatic strokes were reported, and there was no systematic cerebral imaging. Therefore, the rate of silent cerebrovascular embolisms was not assessed. However, the impact of these silent microembolisms remains to be determined.
Among peripheral vascular TAVR, after a pre-specified, propensity-based matching, n-FP and FP TAVR provided similar results and a similar safety profile, except for a 2-fold lower rate of major vascular complications or unplanned vascular repairs in the n-FP TAVR group. Although FP access remains the first choice in TAVR, n-FP TAVR may be a safe alternative when femoral access risk is considered too high and may be favored over surgery in patients who are deemed ineligible for FP TAVR.
COMPETENCY IN PATIENT CARE AND PROCEDURAL SKILLS: n-FP arterial access for TAVR is associated with a lower risk of major vascular complications than femoral access and otherwise similar procedural outcomes.
TRANSLATIONAL OUTLOOK: The outcomes of patients with aortic stenosis who undergo TAVR with nonfemoral peripheral access should be compared with those managed with surgical aortic valve replacement.
Edwards Lifesciences and Medtronic partly funded the FRANCE TAVI registry. Dr. Karam has received consultant fees from Abbott. Dr. Modine has been a consultant for Abbott, Edwards Lifesciences, Medtronic, and Microport. Dr Commeau has received consultant fees from Edwards, Abbott, and Boston Scientific. Dr. Koning has clinical research relationships with Boehringer Ingelheim, Boston Scientific, Abbott, Biosensor, and Biotronik. Dr. Seitz is a shareholder in Volta Medical; and has received consulting fees from Biosense Webster and Abbott. Dr. Spaulding has been a consultant for Medtronic. Dr. Spaulding has received consulting fees from Abiomed, Zoll, Medtronic, and Medpass. Dr. Lefevre has served as a proctor for Edwards Lifesciences and Abbott. Dr. Eltchaninoff has served as a proctor for and received lecture fees from Edwards Lifesciences. Dr. Iung has been a consultant for Edwards Lifesciences; and has received speaker fees from Boehringer Ingelheim and Novartis. Dr. Leprince has served as a proctor for Medtronic and Edwards. Dr. Le Breton has received speaker fees from Edwards Lifesciences and Medtronic. Dr. Lafont has served as a proctor for Medtronic. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- femoral peripheral
- confidence interval
- nonfemoral peripheral
- odds ratio
- transcatheter aortic valve replacement
- Received May 17, 2019.
- Revision received September 12, 2019.
- Accepted September 15, 2019.
- 2019 The Authors
- Ludman P.F.,
- Moat N.,
- de Belder M.A.,
- et al.
- Auffret V.,
- Lefevre T.,
- Van Belle E.,
- et al.
- Patel J.S.,
- Krishnaswamy A.,
- Svensson L.G.,
- Tuzcu E.M.,
- Mick S.,
- Kapadia S.R.
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.
- Folliguet T.,
- Laurent N.,
- Bertram M.,
- et al.
- Amat-Santos I.J.,
- Rojas P.,
- Gutiérrez H.,
- et al.
- Blackstone E.H.,
- Suri R.M.,
- Rajeswaran J.,
- et al.
- Chollet T.,
- Marcheix B.,
- Boudou N.,
- et al.
- Erlebach M.,
- Head S.J.,
- Mylotte D.,
- et al.
- Popma J.J.,
- Deeb G.M.,
- Yakubov S.J.,
- et al.
- Chamandi C.,
- Abi-Akar R.,
- Rodés-Cabau J.,
- et al.
- Husser O.,
- Fujita B.,
- Hengstenberg C.,
- et al.
- Droppa M.,
- Borst O.,
- Katzenberger T.,
- et al.
- Généreux P.,
- Webb J.G.,
- Svensson L.G.,
- et al.
- Rodés-Cabau J.,
- Dumont E.,
- Boone R.H.,
- et al.
- Pagnesi M.,
- Martino E.A.,
- Chiarito M.,
- et al.
- Jochheim D.,
- Zadrozny M.,
- Ricard I.,
- et al.