Transcatheter aortic valve implantation (TAVI) for the treatment of symptomatic and severe aortic valve stenosis (AS) has rapidly evolved during the last decade. TAVI has proven superior or non-inferior against surgical aortic valve replacement (SAVR) for patients at high, intermediate or low surgical risk [1, 2]. Because of superior results on procedural and clinical outcome, the transfemoral technique has been the preferred access for TAVI as compared to transapical access [3, 4, 5]. Safe application of transfemoral access for TAVI is, however, precluded in patients with underlying obstructive peripheral atherosclerotic disease and/or tortuosity of the iliofemoral route (Figs. 1,2). Alternative, non-femoral and non-transapical access approaches for TAVI have thus been developed, such as a transapical, transcaval, direct aortic, transcarotid or transaxillary approach. In recent years, the transaxillary approach has gained popularity in favour of other alternative access sites [6]. This review aims to provide a summary of data available on TAVI performed through transaxillary access.

Fig. 1.

Computed tomography images of iliofemoral access. Stretched vessel views of the right and left iliofemoral artery (A,D) demonstrating advanced obstructive atherosclerotic disease with size of 6mm/18F minimal diameter bar set on the vessel diameter profile (B,E). En face view (C) of the iliofemoral arteries and aortic bifurcation and cross sections of most severe narrowing, with minimal lumen diameter of 3.6 and 2.9 in the right and left femoral artery, respectively.

Fig. 2.

Computed tomography images in preparation for a transaxillary TAVI procedure. (A) Computational rendered image of aortic arch, brachiocephalic trunk and left subclavian artery. (B) Calcifications of the aortic root, arch and ascending aorta including brachiocephalic trunk and left subclavian artery. The calcification is mostly pronounced in the aortic arch and at the ostium of the left subclavian artery with cross-sectional image showing a minimal lumen diameter of 10.9 mm at this level, which is not impeding a trans-axillary access. TAVI, transcatheter aortic valve implantation.

We conducted a review of the literature concerning transaxillary access as alternative route for TAVI. We conclude the following: (1) over the recent years, a wide application of the transaxillary approach for TAVI has developed; (2) this technique has an excellent procedural success rate up to 92%, with low procedural complication rates; and (3) clinical outcomes of transaxillary TAVI appear to be comparable to other alternative access approaches for TAVI.

Data from the STS and American College of Cardiology (ACC) Transcatheter Valve Therapy (TVT) Registry, a large database of TAVI procedures performed in the United States nicely demonstrate the evolution of TAVI access over the years. Transfemoral access was only being used in 57.1% of TAVI cases in 2013 [6, 25]. In the early TAVI years, when the insertion profiles of transcatheter heart valves (THV) were still much larger, a significant amount of procedures were performed via the—Food and Drug Administration (FDA) approved—transapical access reaching up to 34.2% in 2013, and the alternative direct aortic access technique increased up to 8.7% in 2014. However, in the following years, unfavourable data on transapical access, the availability of next-generation devices with lower insertion profiles, and development of other, less invasive, alternative access techniques, has led to an enormous downfall of transapical and direct aortic TAVI procedures (0.5% and 0.3% in 2019, respectively), while overall transfemoral approach increased up to 95.3% of TAVI cases in 2019 [25]. During these years, the transaxillary approach has shown a remarkable rise in popularity. On the other hand, the transcaval access became a newly adopted technique, albeit with limited use (121 cases in 2019), while the transcarotid approach became less common. Transaxillary access became the most commonly employed alternative access technique, accounting for 1816 cases, or 2.5% of all TAVI cases performed in 2019 in the United States [6, 25].

The population of patients who are selected to undergo a TAVI through transaxillary access are generally relatively old with multiple comorbidities and STS-scores ranging between intermediate to high surgical risk. Despite these comorbidities, which are often associated with higher morbidity and mortality after TAVI, procedural success and post-procedural outcomes of transaxillary TAVI appear to be favourable with procedural success of 92% and low post-procedural complications such as MI and life-threatening bleeding [3]. Also, 30-day and 1-year mortality rates of 5% and 17%, respectively, are in line with results of the TVT registry data [25]. On the other hand, major bleeding and vascular complications are more common in transaxillary access cases, often due to closure device failure and/or difficulty compressing the axillary artery. Importantly, these latter do not seem to be associated with worse clinical outcome and are often manageable with percutaneous techniques such as the use of covered stents.

The number of patients suffering from stroke after transaxillary TAVI appears to be low based on our collected data, however, according to larger registries such as the STS-ACC TVT Registry report stroke rates up to 6% after transaxillary TAVI [6]. These results are probably caused by the relatively large amount of studies with small study population in our manuscript. After all, 1 in 3 included studies report on a study population smaller than 30 patients, none of which suffer from post-procedural stroke [7, 9, 10, 11, 12, 14]. Therefore, a caveat must be made concerning post-procedural stroke rates after transaxillary TAVI, as future, larger registries may show higher occurrence of cerebrovascular accident (CVA).

The most frequently occurring VARC-defined outcome in this analysis was the new PPM implantation rate, which reached up to 18% when all data was combined. This higher PPM rate is most probably attributable to the use of mainly first-generation THV devices, that were none repositionable, and were implanted without the adoption of contemporary implantation techniques (e.g., cusp overlap view) that are developed to aim for high implant to reduce the risk for new PPM.

No significant differences in procedural success, 30-day and 1-year mortality were demonstrated in the different studies comparing transaxillary and transfemoral access in this review [16, 17, 18, 19, 20]. The same can be found for other procedural outcomes such as stroke, life-threatening and major bleeding, major vascular complications and AKI with only one of the included studies reporting a significantly higher incidence of MI. Six meta-analyses that compare transaxillary versus transfemoral access are available [3, 4, 26, 27, 28, 29]. In neither of these meta-analyses, a significant difference in 30-day mortality was shown. No significant difference was observed for mortality at 1 year and 1,6 years respectively in two available meta-analyses that report on mid-term results [4, 28]. A third, however, reported a higher midterm mortality (period unspecified) in transaxillary compared to transfemoral cases [3]. This may be explained by the significantly higher midterm stroke rate in the transaxillary group reported in this meta-analysis. Contrarily, in none of the other 5 meta-analyses a difference could be found for stroke rates between transaxillary and transfemoral TAVI. Furthermore there were no significant differences found in all bleeding events in 5 of the meta-analyses [3, 4, 26, 28, 29] while in 2 meta-analyses a significant lower vascular complication rate was seen in the transaxillary group as compared to the transfemoral group [26, 27]. In the remaining 4 meta-analyses there was no difference in vascular complications noted [3, 4, 28, 29]. Fewer AKI was seen in the transaxillary group compared to the transfemoral approach in two of these meta-analysis [3, 29]. The reason for this is unclear, but one hypothesis is that a more direct THV positioning (less ‘slack’) in the annulus due to the shorter route with the transaxillary access might lead to less contrast use.

Whether to decide if transaxillary TAVI is superior to any of the other alternative access techniques, can only be based on data from observational registries, matched cohort retrospective studies and meta-analyses. Up to now, no studies have investigated alternative access techniques in a randomized manner. Patients—with typically a high burden of peripheral atherosclerotic disease—who are not eligible for transfemoral TAVI, often qualify for only limited number of the possible alternative approaches. Furthermore, due to a learning curve aspect, with improvement of outcomes when experience increases, operators tend to gain experience in a limited amount of alternative access approaches. Both imply that the set-up of randomized trials for this matter remains difficult.

In this review, only observational studies were included because of the absence of randomised controlled trials in the area. Furthermore, patient characteristics may differ between the groups, making real head-to-head comparison difficult.

Even though we only included manuscripts using VARC-definitions, the comparison of the results and outcomes was hampered by the omitting of certain endpoints. Moreover, regular revisions of the VARC-definitions—with a recent renewal in VARC-3—complicates the comparison of results across different time periods.

Transaxillary access is considered a feasible and safe alternative approach for TAVI in patients not eligible for implantation of a transcatheter aortic valve by transfemoral access. With increasing experience and subsequent refinement of the technique, transaxillary access may gain further evidence and popularity so that it may further solidify its position as the first-choice alternative to transfemoral TAVI emplacement.

AH—literature search, article reviewing, data extraction, major contributions to writing the article. DTS—literature search, article reviewing, data extraction, major contributions to writing the article. HEJ, MR—major contributions to writing this article, general oversight and supervision, delivered substantial contributions to the conception or design of the work and also played an important role in interpretation of data for the work. LR—major contributions to writing this article, general oversight and supervision, corresponding author, delivered substantial contributions to the conception or design of the work and also played an important role in interpretation of data for the work. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.