Secondary Femoral Access Hemostasis During Transcatheter Aortic Valve Replacement: Impact of Vascular Closure Devices

Background. Vascular and bleeding complications related to secondary femoral access site are frequent in patients undergoing transcatheter aortic valve replacement (TAVR), and their occurrence is associated to poorer outcomes. We aimed to evaluate the clinical impact of vascular closure devices (VCDs) for secondary femoral access hemostasis in TAVR procedures. Methods. This was a multicenter study including 4031 patients who underwent TAVR (mean age, 81 ± 8 years; mean Society of Thoracic Surgeons [STS] score, 4.9 [interquartile range, 3.3-7.6]), and had a secondary femoral access. The 30-day clinical outcomes were analyzed according to femoral access-site hemostasis (manual compression vs VCD), and according to the type of VCD (Perclose [Abbott Cardiovascular] vs Angio-Seal [Terumo Interventional Systems]) using a propensity-matched, multivariable, logistic regression model. Results. Manual compression was used in 941 patients (23.3%) and VCDs were used in 3090 patients (76.7%; Perclose in 1549 patients [38.4%] and Angio-Seal in 1541 patients [38.2%]) for secondary femoral access hemostasis. Vascular complications related to secondary access site occurred in 162 patients (4%), and were more frequent in patients who underwent manual compression (7.2%) compared with VCD hemostasis (3%; adjusted P<.001). In the VCD group, the use of Angio-Seal (vs Perclose) was associated with a higher rate of vascular complications (3.7% vs 2.4%, respectively; adjusted P=.02), femoral artery pseudoaneurysm (1.3% vs 0.4%, respectively; adjusted P<.01), invasive treatment requirement for treating vascular complications (surgery: 0.8% vs 0.3%, respectively [adjusted P=.03]; and thrombin injection: 0.9% vs 0%, respectively [adjusted P<.001]). Conclusion. VCDs represented a safer and more effective alternative compared with manual compression for secondary femoral access-site hemostasis in patients undergoing TAVR procedures, and the Perclose VCD was associated with the lowest risk of vascular complications. Future randomized studies are warranted. 

J INVASIVE CARDIOL 2021;33(8):E604-E613. 

Key words: secondary access, transcatheter aortic valve replacement, vascular closure device, vascular complication

Vascular and bleeding complications are common in patients undergoing transcatheter aortic valve replacement (TAVR).¹ Their occurrence has been associated with an increased risk of short- and long-term mortality, as well as with a longer hospital stay and increased readmission rate.^2,3 The contralateral femoral artery remains the most common secondary access in TAVR procedures, and vascular complications related to the secondary femoral access site account for up to one-fourth of global vascular complications in transfemoral TAVR recipients.^4-6 Additionally, major vascular complications related to the secondary femoral access have been associated with a similar negative clinical impact and increased mortality as those related to the primary access in TAVR procedures.⁴ 

Nowadays, vascular closure devices (VCDs) are widely used to achieve femoral hemostasis after procedures requiring common femoral artery puncture, and their use has been associated with a reduced time to hemostasis and earlier ambulation compared with manual compression (MC).^7-10 Nonetheless, there is no consensus on whether VCDs also reduce the incidence of vascular and bleeding complications.^7-13 In the TAVR field, the potential beneficial effects of VCDs (such as a shorter time to hemostasis or to ambulation) are limited, as TAVR recipients usually rest in bed and stay in hospital for longer periods of time compared with those undergoing coronary angiography or percutaneous coronary intervention. Consequently, the main advantage of using a VCD for femoral secondary access-site hemostasis in TAVR procedures would be the reduction of the risk of vascular complications. However, no data exist on the safety and efficacy of VCDs for femoral secondary access hemostasis in the TAVR population. Additionally, the current TAVR population is elderly and exhibits a high prevalence of comorbidities, and to date, most VCD studies have excluded patients at high risk for vascular complications. Finally, no direct comparative studies between different VCDs for secondary access hemostasis exist in this population. Thus, the objectives of the present study were: (1) to determine the occurrence of vascular and bleeding events related to femoral secondary access site, according to femoral hemostasis (VCD vs MC), in a large cohort of patients undergoing TAVR; and (2) to compare the incidence of vascular and bleeding complication according to two different VCDs (Perclose [Abbott Cardiovascular] vs Angio-Seal [Terumo Interventional Systems). 

A total of 5096 patients who underwent TAVR in 11 tertiary-care centers in Europe and Canada between 2007 and 2019 were included. Of these, 1065 patients were excluded as secondary access was performed via the radial artery, leading to a final study population of 4031 patients. Primary access for TAVR procedures was transfemoral in 3298 patients (81.8%), transapical in 418 patients (10.4%), transcarotid in 135 patients (3.4%), trans-subclavian in 93 patients (2.3%), transaortic in 81 patients (2.0%), and transcaval in 6 patients (0.2%). The selection of the femoral artery as secondary access and the hemostasis method were left to the discretion of the heart team responsible for the patient/procedure.

Baseline, periprocedural and 30-day clinical event data were prospectively collected in a dedicated database at each participating center. All events were defined according to the Valve Academic Research Consortium-2 criteria.¹⁴ Data collection for this study was performed in accordance with the ethics committee of each participating center and all patients provided signed informed consent for the procedures.

Secondary femoral access and hemostasis. The common femoral artery was identified by palpation; occasionally, fluoroscopy was used to locate the femoral head. Afterward, a 5 Fr (14.3%), 6 Fr (70.6%), or 7 Fr sheath (15.1%) was used. Ultrasound-guided access was not performed. In patients who underwent MC, the sheath was removed by a physician and MC was applied proximal to the puncture site. Compression was continued until hemostasis was achieved. The VCDs were deployed according to standard practice. Two different devices were used in this study: the suture-based Perclose Proglide VCD and the collagen-based Angio-Seal VCD. A pressure bandage after hemostasis was not applied systematically. Patients were required to rest in bed for 4 to 6 hours, and if no vascular or bleeding complications were identified during this time, ambulation was initiated.  

Statistical analysis. Continuous variables were presented as mean ± standard deviation and compared using the t-test or Wilcoxon rank-sum test. Categorical variables were presented as counts and frequency and compared using the Chi-square or Fisher’s exact test as appropriate. The unadjusted analyses were initially performed in the overall study population, according to the type of hemostasis (MC vs VCD). A propensity-score analysis was performed to adjust for the intergroup clinical differences. A propensity score representing the likelihood of having a VCD was calculated for each patient by the use of multivariable logistic regression analysis that identified variables independently associated with the use of a VCD at a P<.20. Continuous variables were checked for the assumption of linearity in the logit and the graphical representations suggested linear relationships. Interaction between variables were allowed only if it was supported clinically and statistically (P<.20). Variables retained in the final model were hypertension, previous coronary artery disease, diabetes, peripheral vascular disease, kidney disease, type of valve, and center. Afterward, a second analysis including only the patients who received a VCD was performed. The unadjusted analyses included the cases in which a VCD was used, according to the type of device (Perclose vs Angio-Seal). A propensity score representing the likelihood of having each device was calculated for each patient by the use of multivariable logistic regression analysis (similar to the one described above). Variables retained in the final model were gender, body mass index, hypertension, diabetes, kidney disease, atrial fibrillation, peripheral vascular disease, type of valve, and center. All analyses were performed using a hierarchical method to account for between-center variability. The 30-day clinical outcomes were compared between groups with the use of logistic regression hazard models. The propensity score was then incorporated as an additional variable in the logistic regression models. Finally, a logistic regression was performed to determine independent predictors of vascular complications. Those variables from the univariable analysis with a P-value <.10 were entered into a multivariable regression analysis. The results were considered significant if P was <.05. All analyses were conducted using the statistical package SAS, version 9.4 (SAS Institute).

Manual compression vs vascular closure device. The main baseline clinical and procedural characteristics of the overall population and according to the type of femoral hemostasis (MC vs VCD) are shown in Table 1. The mean age of the study population was 81 ± 8 years, and 52% were women. Patients in the MC group had a higher prevalence of diabetes mellitus (36.8% vs 31.2%; P<.01), hypertension (87.0% vs 77.1%; P<.001), coronary artery disease (58.3% vs 38.7%; P<.001), and peripheral artery disease (27.5% vs 15.8%; P<.001), and exhibited an increased surgical risk (Society of Thoracic Surgeons Predicted Risk of Mortality score, 5.6% [interquartile range, 3.7-8.4] vs 4.6 [interquartile range, 3.2-7.2]; P<.001).

The 30-day clinical outcomes of the overall population and according to the type of secondary access hemostasis (MC or VCD) are shown in Table 2. Vascular complications related to secondary access site were observed in 68 patients (7.2%) in the MC group vs 94 patients (3.0%) in the VCD group (P<.001). This difference was mainly driven by a higher incidence of minor vascular complications in the MC group (5.2% in the MC group vs 1.7% in the VCD group; P<.001), whereas the rate of major vascular complications was numerically higher (but not statistically significant) in the MC group (2.0% in the MC group vs 1.3% in the VCD group; P=.17). In the multivariate analysis (Table 3), the type of femoral hemostasis was the sole factor associated with the occurrence of secondary access-site vascular complications (MC: odds ratio [OR], 2.03; 95% confidence interval [CI], 1.42-2.92; P<.001). A detailed description of the vascular complications is provided in Figure 1 and Table 4. Femoral hematoma was the most frequent vascular complication in both groups, accounting for 48.5% and 48.9% of the vascular complications in the MC and VCD hemostasis groups, respectively. Up to 3.1% of the patients who underwent MC developed a pseudoaneurysm, in comparison with 0.8% in the VCD group (P<.001); consequently, a higher number of patients needed thrombin injection to treat this complication in the MC group (1.5% in the MC group vs 0.5% in the VCD group; P<.001). 

Regarding other 30-day outcomes, the rate of major/life-threatening bleeding related to secondary femoral access site was comparable between groups (1.6% in the MC group vs 1.0% in the VCD group; P=.26), and the rate of minor bleeding was higher in the MC group (2.8% in the MC group vs 1.3% in the VCD group; P<.01). Finally, acute renal failure and mortality rates were higher in the MC group (17.0% in the MC group vs 10.1% in the VCD group [P<.001] and 6.4% in the MC group vs 3.8% in the VCD group [P<.01], respectively). 

Vascular closure devices: Perclose vs Angio-Seal. The main baseline clinical and procedural characteristics and 30-day events according to the femoral VCD (Perclose vs Angio-Seal) are shown in Table 5 and Table 6, respectively. Vascular complications related to secondary access site were more frequent in the group of patients who received the collagen-based Angio-Seal device (3.7% in the Angio-Seal group vs 2.4% in the Perclose group; P=.02). The type of vascular complications associated with each device was similar (Figure 2 and Table 7), with the exception of femoral pseudoaneurysms, which were identified in 1.3% of the Angio-Seal group vs 0.4% of the Perclose group (P<.01). Also, vascular complications required different treatments according to the device used (vascular surgery was required in 0.3% of the Perclose group vs 0.8% of the Angio-Seal group [P=.03] and thrombin injection was required in 0% of the Perclose vs 0.9% of the Angio-Seal group [P<.001]). 

There were no differences between groups regarding secondary femoral access-site major/life-threatening bleeding complications (1.0% in the Perclose group vs 1.0% in the Angio-Seal group; P=.85) or minor bleeding complications (1.2% in the Perclose group vs 1.5% in the Angio-Seal group; P=.34), acute renal failure (9.2% in the Perclose group vs 11.1% in the Angio-Seal group; P=.29), or mortality (4.4% in the Perclose group vs 3.1% in the Angio-Seal group; P=.13).  

This is the first study evaluating the clinical outcomes according to the type of secondary femoral access-site hemostasis in the TAVR population. The main findings can be summarized as follows: (1) in patients undergoing TAVR, MC for secondary femoral access-site hemostasis was associated with a higher rate of vascular complications compared with VCDs, and MC was the only factor independently associated with secondary access vascular complications; (2) in patients in whom a VCD was used for secondary access hemostasis, the suture-based Perclose device showed the lowest rate of vascular complications (compared with the collagen-based Angio-Seal device); and (3) the occurrence of femoral pseudoaneurysm was more frequent after MC (compared with VCD), and in those patients who received a VCD, the incidence of pseudoaneurysm was higher with the Angio-Seal device (vs Perclose). 

Manual compression continues to be the standard of care for achieving femoral hemostasis following a femoral puncture in many centers, as it is identified as a simple, safe, and inexpensive method.¹¹ Nonetheless, MC could also be time consuming for the operator and uncomfortable for the patient, and usually requires patient prolonged bedrest. In the 1990s, VCDs were introduced in the clinical practice to facilitate femoral hemostasis.^10,15 The principal reasons for using VCDs were to reduce time to hemostasis, facilitate earlier mobilization after sheath removal, and enhance patient comfort.^10,15 While some studies (including patients who underwent coronary and peripheral interventions) showed improved or comparable results with VCD vs MC, others suggested an increased risk of vascular complications associated with the use of VCDs.^7-13 Nonetheless, a variety of VCD options were included in these studies, and most of them were underpowered to show differences between groups.^7-13

In TAVR procedures, the well-established advantages of VCDs (patient comfort and early mobilization) may not be as useful as in coronary cases, as bedrest for a longer period of time is usually required after TAVR. Additionally, the advantages regarding same-day discharge — and, consequently, lower hospitalization costs — would not be applicable to the TAVR population.¹³ Thus, the main reason for using a VCD to facilitate secondary femoral access hemostasis in TAVR procedures would be to reduce the rate of vascular complications. In the TAVR field, primary or secondary access-site related vascular complications and bleeding events have been associated with poorer outcomes, including higher mortality, more prolonged hospitalizations, and increased costs.^3,4 Even though VCDs have failed to demonstrate a real advantage in coronary or peripheral interventions, results from our study showed that they seem to be a safer option compared with MC in TAVR recipients, reducing the global rate of vascular complications. Additionally, in our cohort, MC was identified as the single factor independently associated with secondary femoral access-site vascular complications. Unlike previous studies in the coronary/peripheral field, our study is the first to evaluate femoral hemostasis in an all-comers study including TAVR recipients. Most of the randomized and observational studies that have previously compared VCD with MC excluded patients at high risk of puncture-site complications, including lower-limb atherosclerosis, calcified or small common femoral arteries, and obesity, characteristics frequently present in the TAVR population.⁷ Nonetheless, a greater benefit with the use of a VCD in high-risk patients undergoing percutaneous coronary interventions has also been described.¹³ 

In line with some previous studies in the coronary/peripheral field, we found that the occurrence of hematoma (classified as a major or minor vascular complication) and pseudoaneurysm were significantly lower if a VCD was used.^9,13,16,17 Pseudoaneurysm is one of the most feared local access-site complications, as its occurrence usually requires either percutaneous or surgical repair. Our study showed that the incidence of femoral pseudoaneurysm was significantly higher in the MC group vs the VCD group, despite the use of smaller sheath sizes in the MC group. Previously, Naddaf et al¹⁸ identified the lack of VCD use as the strongest predictor for the occurrence of femoral pseudoaneurysm, and Kerré et al¹³ also found a reduction in the incidence of femoral pseudoaneurysm if a VCD was used. 

The 2 VCDs used in this study (Perclose and Angio-Seal) have been previously associated with a significant decrease in time to hemostasis and time to ambulation, as well as a comparable/lower rate of vascular complications compared with MC.^19,20 However, despite the fact that similar rates of vascular complications have been previously reported for the collagen-based Angio-Seal in comparison with the suture-based Perclose VCD,²¹ these two devices have not been previously compared in a TAVR population. Interestingly, in our study, the rate of vascular complications, including femoral pseudoaneurysm, was higher in patients who underwent hemostasis with the Angio-Seal device vs those who received a Perclose device. Previous studies showed an incidence of femoral pseudoaneurysm up to 3% with the use of the Angio-Seal device, while very few cases of pseudoaneurysm have been reported after femoral hemostasis with the Perclose device.^13,22-25 Additionally, higher rates of VCD failure have been reported for the suture-based Perclose device.^21,26,27 In our study, the rates of VCD failure were low and similar between the two devices, probably as a consequence of the wide expertise of TAVR operators for the use of the Perclose device. This low rate of VCD failure highlights the safety of using a VCD despite the high-risk profile of the TAVR population. 

Study limitations. First, this study was subject to the limitations inherent in non-randomized studies and retrospective data analyses. However, a propensity-score matched analysis was performed to adjust for baseline/procedural differences, and all complications were prospectively collected and entered in a dedicated database. Second, time to hemostasis and time to ambulation were not recorded. However, time to ambulation in TAVR patients would be mainly driven by primary access hemostasis, with minimal or no impact from secondary access hemostasis.

VCDs seem to offer a safer alternative to MC for achieving secondary femoral access-site hemostasis in patients undergoing TAVR procedures. The suture-based Perclose device was associated with the lowest incidence of vascular complications, and the collagen-based Angio-Seal device accounted for a higher rate of femoral pseudoaneurysm and required more frequent invasive treatments to treat vascular complications. Future randomized trials are warranted. 

From the ¹Quebec Heart and Lung Institute, Laval University, Quebec City, Canada; ²Assistance Publique-Hôpitaux de Paris, Bichat Hospital, Paris, France; ³Hospital Universitario Virgen de la Victoria, Málaga, Spain; ⁴Instituto Cardiovascular, Hospital Clinico San Carlos, IdISSC, Madrid, Spain; ⁵Groupe Hospitalier Mutualiste de Grenoble, Institut Cardiovasculaire, Grenoble, France; ⁶Hospital Marques de Valdecilla, Santander, Spain; ⁷Hospital Universitario Central de Asturias, Oviedo, Spain; ⁸Hospital Universitari Vall d'Hebron, Barcelona, Spain; and ⁹Hospital Clínic de Barcelona, Barcelona, Spain.

Funding: Dr. Lucía Junquera was supported by a research grant from the Fundación Alfonso Martín Escudero (Madrid, Spain). Dr Josep Rodés-Cabau holds the Research Chair “Fondation Famille Jacques Larivière” for the Development of Structural Heart Disease Interventions. 

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript accepted November 13, 2020.

Address for correspondence: Josep Rodés-Cabau, MD, Quebec Heart and Lung Institute, Laval University, 2725 Ch Ste-Foy, Québec, QC G1V 4G5, Canada. Email: josep.rodes@criucpq.ulaval.ca