Vascular Complications After Transcatheter Aortic Valve ReplacementInsights From the PARTNER (Placement of AoRTic TraNscathetER Valve) Trial
Author + information
- Philippe Généreux, MD⁎,
- John G. Webb, MD†,
- Lars G. Svensson, MD, PhD‡,
- Susheel K. Kodali, MD⁎,
- Lowell F. Satler, MD§,
- William F. Fearon, MD∥,
- Charles J. Davidson, MD¶,
- Andrew C. Eisenhauer, MD#,
- Raj R. Makkar, MD⁎⁎,
- Geoffrey W. Bergman, MB, BS††,
- Vasilis Babaliaros, MD‡‡,
- Joseph E. Bavaria, MD§§,
- Omaida C. Velazquez, MD∥∥,
- Mathew R. Williams, MD⁎,
- Irene Hueter, PhD¶¶,
- Ke Xu, PhD⁎,
- Martin B. Leon, MD⁎,⁎ (mleon{at}crf.org),
- PARTNER Trial Investigators
- ↵⁎Reprint requests and correspondence:
Dr. Martin B. Leon, Columbia University Medical Center, New York-Presbyterian Hospital, 161 Fort Washington Avenue, 6th Floor, New York, New York 10032
Abstract
Objectives This study sought to identify incidence, predictors, and impact of vascular complications (VC) after transfemoral (TF) transcatheter aortic valve replacement (TAVR).
Background VC after TF-TAVR are frequent and may be associated with unfavorable prognosis.
Methods From the randomized controlled PARTNER (Placement of AoRTic TraNscathetER Valve) trial, a total of 419 patients (177 from cohort B [inoperable] and 242 from cohort A [operable high-risk]) were randomly assigned to TF-TAVR and actually received the designated treatment. First-generation Edwards-Sapien valves and delivery systems were used, via a 22- or 24-F sheath. The 30-day rates of major and minor VC (modified Valve Academic Research Consortium definitions), predictors, and effect on 1-year mortality were assessed.
Results Sixty-four patients (15.3%) had major VC and 50 patients (11.9%) had minor VC within 30 days of the procedure. Among patients with major VC, vascular dissection (62.8%), perforation (31.3%), and access-site hematoma (22.9%) were the most frequent modes of presentation. Major VC, but not minor VC, were associated with significantly higher 30-day rates of major bleeding, transfusions, and renal failure requiring dialysis, and with a significantly higher rate of 30-day and 1-year mortality. The only identifiable independent predictor of major VC was female gender (hazard ratio [HR]: 2.31 [95% confidence interval (CI): 1.08 to 4.98], p = 0.03). Major VC (HR: 2.31 [95% CI: 1.20 to 4.43], p = 0.012), and renal disease at baseline (HR: 2.26 [95% CI: 1.34 to 3.81], p = 0.002) were identified as independent predictors of 1-year mortality.
Conclusions Major VC were frequent after TF-TAVR in the PARTNER trial using first-generation devices and were associated with high mortality. However, the incidence and impact of major VC on 1-year mortality decreased with lower-risk populations.
Vascular complications (VC) after transfemoral (TF) transcatheter aortic valve replacement (TAVR) are frequent and may be associated with unfavorable clinical outcomes (1–8). The use of large-diameter catheters (18-F to 24-F) and the high-risk characteristics of the population treated in the early days of TAVR may explain the high incidence. Although the PARTNER (Placement of AoRTic TraNscathetER Valve) trial recently demonstrated that TAVR is associated with similar mortality at 30 days and up to 2 years in surgical high-risk patients compared with surgical aortic valve replacement (9,10), and that TAVR is superior to medical treatment in patients not suitable for conventional surgery (11,12), a number of TAVR-associated complications have been identified, including a high incidence of VC. Published studies using the first-generation devices showed an incidence of major VC varying from 5% to 23% using uniform definitions (13,14). Recently published data (7) have suggested improvement in VC, due to the combination of newer device generations, smaller delivery systems, and the use of adjunctive techniques, combined with better screening and increased operator experience (7). The aim of this report is to better characterize the incidence, nature, predictors, and impact of VC on long-term prognosis after TF-TAVR from the multicenter, prospective randomized PARTNER trial.
Methods
Study population
The design and initial results of the PARTNER trial (cohort B and cohort A) have been published previously (9,11). Briefly, the PARTNER trial enrolled patients with severe symptomatic aortic stenosis. Patients were divided into 2 cohorts: those who were considered to be candidates for surgery despite the fact that they were at high surgical risk, as defined by a Society of Thoracic Surgeons risk score of 10% or higher or by the presence of coexisting conditions that would be associated with a predicted risk of death by 30 days after surgery of 15% or higher (cohort A), and those who were not considered to be suitable candidates for surgery because they had coexisting conditions that would be associated with a predicted probability of 50% or more of either death by 30 days after surgery or a serious irreversible condition (cohort B).
Patients from cohort B with a suitable iliofemoral vessel were then randomized to TF-TAVR with the Edwards-Sapien heart valve system (Edwards Lifesciences, Irvine, California) or to standard medical care. Patients enrolled in cohort A were then randomized to TAVR (TF if iliac and femoral vessels were suitable or transapical if not) or to conventional surgical aortic valve replacement. The current analysis pooled patients from cohorts A and B who underwent TAVR via TF approach only. The study was approved by the institutional review board at each participating site, and all patients provided written informed consent.
Study endpoint
VC were defined according to a modified version of the Valve Academic Research Consortium criteria as described in the PARTNER trial protocol (11,14). Major VC were defined by the presence of any of the following: 1) any thoracic aortic dissection; 2) access site or access-related vascular injury (dissection, stenosis, perforation, rupture, arteriovenous fistula, pseudoaneurysm, hematoma, irreversible nerve injury, or compartment syndrome) leading to either death, need for significant blood transfusions (≥4 U), unplanned percutaneous (endovascular stent) or surgical intervention, or irreversible end-organ damage; 3) distal embolization (noncerebral) from a vascular source requiring surgery or resulting in amputation or irreversible end-organ damage; or 4) left ventricular perforation. Minor VC were defined by the presence of any of the following: 1) access site or access-related vascular injury (dissection, stenosis, perforation, rupture, arteriovenous fistula or pseudoaneurysm requiring compression or thrombin injection therapy, or hematomas requiring transfusion (≥2 but <4 U) but not requiring unplanned percutaneous or surgical intervention and not resulting in irreversible end-organ damage; 2) distal embolization treated with embolectomy and/or thrombectomy and not resulting in amputation or irreversible end-organ damage; and 3) failure of percutaneous access site closure resulting in interventional (endovascular stent) or surgical correction and not associated with death, need for significant blood transfusions (≥4 U), or irreversible end-organ damage.
The 30-day and 1-year frequency of all-cause mortality, cardiovascular mortality, stroke, major bleeding, myocardial infarction, and acute kidney injury were reported according to Valve Academic Research Consortium definitions (14). All adverse events were adjudicated by an independent clinical events committee. Independent core laboratories analyzed all echocardiograms and electrocardiograms. All data were sent for analysis to an independent academic biostatistics group.
Statistical analysis
All the analyses were performed with data from the as-treated population, which included all patients who underwent TF-TAVR as the final treatment of the index procedure. Continuous variables are summarized as mean ± SD or medians and quartiles, as appropriate, and were compared using the Student t test or Mann-Whitney rank sum test accordingly. Categorical variables were compared by the chi-square or the Fisher exact test. Survival curves for time-to-event variables were constructed on the basis of all available follow-up data with the use of Kaplan-Meier estimates and comparisons relied on the log-rank test. Multivariate logistic regression was performed to identify independent predictors of 30-day major VC (α = 0.05). The multivariable model was built by selecting variables of clinical interest and/or satisfying the entry criterion of p < 0.05 in the univariate analysis. Variables included in the model were carefully selected to avoid overfitting. All selected variables were entered at the same time. To assess the association between major VC and 1-year rate of all-cause mortality, Cox multivariable regression analyses were performed, with variable selection performed as described in the preceding text. A 2-sided alpha level of 0.05 was used for all superiority testing. All statistical analyses were performed with the use of SAS software, version 9.2 (SAS Institute, Cary, North Carolina).
Results
Patients and baseline characteristics
Among the 699 patients enrolled in the PARTNER trial cohort A and the 358 patients enrolled in the PARTNER trial cohort B, 244 and 179, respectively, were randomized to TF-TAVR from these patients, 242 from cohort A and 177 from cohort B (n = 419) were actually treated via TF access and were included in this analysis. Among them, 64 (15.3%) had a major VC and 50 (11.9%) had a minor VC within 30 days of the index procedure. Baseline and procedural characteristics of patients stratified by occurrence of major VC within 30 days are shown in Tables 1 and 2.⇓ Compared with patients with no major VC, patients with major VC were more frequently female, had a lower body surface area, more frequently had insulin-treated diabetes at baseline, had smaller vessel diameter, and had higher sheath-to-femoral-artery ratio and sheath-to-external-iliac-artery ratio. During the procedure, major VC were associated with higher rates of embolization or migration of the prosthesis, use of hemodynamic support devices, and conversion to open heart surgery. The index procedure was longer in the major VC group, with more contrast used, and longer fluoroscopy time. Access and closure of the access site was performed in 76.8% of patients by surgical cutdown, whereas suture-based closure device systems were used in 24.9% of the complete cohort. Surgical cutdown was used more frequently in the major VC group (85.7% vs. 75.2%, p = 0.07), whereas a complete percutaneous approach using closure devices was used more often in the non-major VC group (15.9% vs. 26.5%, p = 0.1). The duration of hospitalization after the procedure was longer in the major VC group (6.9 ± 2.5 days vs. 5.8 ± 2.3 days, p = 0.006).
Baseline Clinical Characteristics of Patients According to the Occurrence of Major VC Up to 30 Days
Procedural Characteristics According to the Occurrence of VCs Up to 30 Days
Clinical outcomes
Types of major VC and their 30-day rates of occurrence are shown in Figure 1. Vascular dissection, vascular perforation, and access site hematoma were most frequently reported. Three cases (0.7%) of aortic dissection and 2 cases (0.5%) of left ventricle perforation were reported. Table 3 shows strategies used to manage major VC. Thirty-day rates for major, minor, and the different types of VC stratified by PARTNER cohort (cohort B vs. A) are shown in Figure 2.
Distribution of Type of Major Vascular Complications After Transcatheter Aortic Valve Replacement
Vascular dissection, vessel perforation, and access site hematoma were the most frequent causes of major vascular complications.
Difference in Rates of VC From Cohort 1B to Cohort 1A
The rates of all vascular complications (VC), vascular dissection, and access site hematoma decreased from cohort 1B to cohort 1A, suggesting improved outcomes in a lower-risk population and with more experienced operators. TF = transfemoral.
Management Strategies Among 64 Patients With Major VC
Clinical outcomes of patients stratified by major VC versus no major VC are shown in Table 4 and in Figure 3. The occurrence of major VC after TAVR was associated with significantly higher 30-day rates of major bleeding and transfusions, and with significantly higher rates of 30-day and 1-year all-cause and cardiac mortality. Major VC were also associated with a significantly higher rate of renal failure requiring dialysis at 30 days. Clinical outcomes of patients stratified by minor VC versus no vascular complication are shown in Table 5 and Figure 4. At 30 days and 1 year, the occurrence of minor VC was not associated with higher mortality, major bleeding, transfusions, or acute kidney injury requiring dialysis compared with patients with no VC at all.
Kaplan-Meier Curves Showing Cumulative Death Rate Through 1 Year for Major VC
Comparison of the cumulative death rate through 1 year in patients with major VC compared with patients with no major vascular complications (VC). Total population (A), cohort 1B only (B), cohort 1A only (C). The impact of major VC on mortality decreased from cohort 1B (B) to cohort 1A (C). CI = confidence interval; HR = hazard ratio.
Kaplan-Meier Curves Showing Cumulative Death Rate Through 1 Year for Minor VC
Comparison of the cumulative death rate through 1 year between patients with minor VC compared with patients with no VC at all. The occurrence of minor VC did not impact mortality. Abbreviations as in Figure 3.
30-Day and 1-Year Event Rates According to the Occurrence of Major VC Versus No Major VC Up to 30 Days
30-Day and 1-Year Event Rates According to the Occurrence of Minor VC Versus No VC Up to 30 Days
Multivariate analysis
Variables associated with major VC are shown in Table 6. After multivariable analysis, female sex was identified as the strongest independent predictor of major VC (hazard ratio [HR]: 2.31 [95% confidence interval (CI): 1.08 to 4.98], p = 0.03). A multivariate Cox proportional hazard analysis including patients from both cohorts identified major VC (HR: 2.31 [95% CI: 1.20 to 4.43], p = 0.012) and renal disease at baseline (HR: 2.26 [95% CI: 1.34 to 3.81], p = 0.002) as independent predictors of 1-year mortality (Table 7).
Independent Clinical Predictors of Major VCs Within 30 Days
Independent Clinical Predictors of 1-Year Death After TAVR
Discussion
The current report, drawn from a cohort of 419 patients with severe symptomatic aortic stenosis who underwent TF-TAVR, is the largest study to specifically evaluate the incidence, predictors of, and impact of major VC on long-term prognosis. The main results of the present study are as follows: 1) major VC after TF-TAVR using the first generation of large devices were frequent; 2) the occurrence of major VC after TAVR was associated with bleeding events, transfusions, and increased mortality; 3) the incidence and impact of major VC seems to decrease in a lower-risk population. The current study demonstrated the prognostic impact of major VC after TAVR on short- and long-term mortality. Indeed, major VC were associated with a more than 4-fold increase in 30-day mortality. Several other investigators have reported similar results. Rodes-Cabau et al. (15) identified access site complications as a potential factor associated with 30-day and cumulative late mortality (median follow-up of 8 months). Similarly, Thomas et al. (16), in the SOURCE (SAPIEN Aortic Bioprosthesis European Outcome) registry, demonstrate a numerically higher, but nonstatistically significant, rate of death at 30 days in patients with major VC compared with no major VC (12.2% vs. 5.6%, p = 0.108). Ducrocq et al. (4) also showed comparable results, with a 30-day mortality of 11.1% in the VC group compared with 4.5% in the non–VC group (p = 0.42). Despite important differences in outcome definitions and in the adjudication of events between these studies and the PARTNER trial, these findings highlight the prognostic importance of major VC after TAVR.
Not surprisingly, major VC were associated with a high rate of major bleeding (60.9%) and transfusions (40.7%) at 30 days in our cohort. These results parallel the early case series published by Ducrocq et al. (4), in which 78% of patients with VC needed a transfusion. Interestingly, more episodes of acute renal failure requiring dialysis occurred in the major VC group. This finding reflected the severity of some of the major VC that occurred in our cohort, leading to severe hemodynamic collapse and/or embolic events, more transfusions, more contrast used, and eventually, kidney function compromise. Also, more invasive strategies (conversion to open heart surgery, hemodynamic support devices) were used to manage major VC, reflecting the severity of the initial vascular insult and may, per se, be involved in the genesis of significant kidney injury. Conversely, the occurrence of minor VC had no impact on short- and long-term mortality, and was associated with less severe bleeding, fewer transfusions, and a lower rate of acute kidney injury requiring renal therapy support.
Important predictors of major VC after TAVR have been identified by multiple groups. Early experience of the site and operators, severe femoral artery calcification, minimal artery diameter, and a sheath-to femoral-artery ratio >1.05 have been associated with the occurrence of VC (7,8). In the current analysis, female sex was the only identifiable independent predictor of major VC after TAVR. The smaller vessel diameters encountered in women paired with the relatively large first-generation introducer-sheath catheter system used in the PARTNER trial (22-F or 24-F) may partially explain this finding. This presumption has been confirmed recently by Hayashida et al. (17), in which women had smaller minimal femoral sizes, higher sheath-to-artery ratios, and were more likely to experience iliac complications after TAVR compared with men. However, even after adjusting for sheath-to-femoral-artery ratio, female sex remained a strong independent predictor of major VC after TAVR, implying that impact of sex goes beyond these simplistic explanations. An intrinsic increased vulnerability to periprocedural complication during and early after an invasive procedure may be present in women. Exact mechanisms for this finding remain elusive and warrant further investigation. Interestingly, our result suggests that although female sex is a strong predictor of periprocedural (vascular) complications (Table 6), at 1 year, female sex is associated with a reduction in all-cause mortality (Table 7). This finding has also been reported by another group (17). Thus, the impact of sex on short- and long-term outcomes is complex, and a more detailed analysis is needed.
Full percutaneous TAVR with the use of closure devices was performed in only 24.9% of patients in the combined population of cohorts B and A. This low rate, in contrast with contemporary practice, may be explained by the fact that many centers among the PARTNER trial sites were at the beginning of their learning curve. Also, current closure devices (Prostar XL and PerClose ProGlide, both Abbott Vascular, Santa Clara, California) are only approved for closure of ≤10-F sheath holes in the United States. Access and closure via surgical cutdown may have been viewed as more predictable, offering more direct control during adverse events, especially in less experienced hands. Although no studies have specifically compared both approaches, recent data suggest that a full percutaneous procedure, helped by adjunctive techniques, can be performed with a low rate of VC and bleeding events, especially with the use of a smaller diameter sheath (5–7). In addition, techniques such as the “cross-over balloon occlusion technique,” in which an endovascular balloon, brought either via the contralateral femoral access site or via the radial artery, is inflated proximal to the site of arteriotomy prior to the large sheath removal, has allowed safer execution of a fully percutaneous TAVR procedure (5,6,18). Although percutaneous closure has been routinely adopted by many centers and performed with a high rate of success after TAVR, a learning curve and specific complications should be acknowledged (6,19). The superiority of one approach (e.g., surgical cutdown) compared with the other (e.g., closure device) still remains a matter of debate, and outcomes are also expected to vary according to the expertise of operators.
Several authors have already demonstrated the importance of increased experience and improved device systems (20–22). Toggweiler et al. (7) recently demonstrated the combined effect of increased experience paired with smaller sheath size use on the occurrence of VC. The rate of major VC went from 8.0% in 2009, using mainly a first-generation 22- to 24-F sheath system, to 1% in 2010, using a smaller 18- to 19-F sheath system. From the PARTNER trial experience, VC not only decreased in cohort A patients (high-risk but operable) as compared with cohort B patients (nonoperable), but also had less impact on short- and long-term mortality (Fig. 3C). Many factors could explain these findings: 1) temporally speaking, cohort B was one of the first TAVR studies to be performed in the United States, and all enrolling sites/operators were in the early stages of experience, learning about appropriate patient selection, vascular screening, and crucial technical details of the procedure; 2) patients from cohort B, being sicker, were more fragile and less inclined to tolerate any major complications; and 3) the transapical approach became available in cohort A, in cases where the iliofemoral vessels were not suitable for the TF approach, making less likely the scenario of “borderline” vessel diameter or vascular anatomy being attempted by the TF approach. These findings underline the magnitude of impact and the synergy between enhanced operator skills, TAVR team experience, and device improvement. These 3 factors are vital if the full clinical potential of the TAVR procedure is to be realized, especially when involving a lower-risk population.
Study limitations
As an observational post hoc analysis, it can only identify correlations, not prove causality. Despite adjustment for potential confounders, unmeasured variables may not have been fully controlled. All these findings, therefore, should be considered hypothesis generating. Only suitable iliofemoral vessel access was included in the PARTNER trial and considered for TF-TAVR. Although rigorous vascular screening, including angiograms, computed tomography scans, and/or peripheral intravascular ultrasound, was performed before randomization, no core laboratory analysis or prospective data collection of these imaging studies was included in case report forms, and all analyses were performed retrospectively with available imaging data. Finally, as mentioned, the PARTNER trial was performed using the first-generation devices with operators and sites at the beginning of their learning curve.
Conclusions
Major VC were frequent after TF-TAVR in the PARTNER trial using first-generation devices and were associated with a high 1-year mortality. However, the incidence and the impact of major VC decreased with lower-risk populations.
Acknowledgments
The authors would like to thank Viral Gandhi (Edwards Lifesciences) and Maria Alu (Columbia University Medical Center) for their important contribution in additional data extraction and final revision of the manuscript.
Footnotes
The PARTNER trial was funded by Edwards Lifesciences and designed collaboratively by the Steering Committee and the sponsor. The present analysis was carried out by academic investigators at the Cardiovascular Research Foundation. Dr. Généreux has received speaker honoraria, consulting fees, and research grant from Edwards Lifesciences. Dr. Webb has received consulting fees from Edwards Lifesciences, as well as travel reimbursement for activities related to his participation on the Executive Committee of the PARTNER trial. Dr. Svensson has received travel reimbursement from Edwards Lifesciences for activities related to his participation on the Executive Committee of the PARTNER trial. Dr. Kodali has received consulting fees from Edwards Lifesciences and Medtronic; and is a member of the Scientific Advisory Board of Thubrikar Aortic Valve, Inc., the Medical Advisory Board of Paieon Medical, and the TAVI Advisory Board of St. Jude Medical. Dr. Davidson received grant support and consulting fees from Edwards Lifesciences. Dr. Fearon has received travel reimbursement from Edwards Lifesciences for activities related to his participation on the Steering Committee of the PARTNER 2 trial. Dr. Makkar has received grant support and travel reimbursements from Edwards Lifesciences; consulting fees from Cordis, Medtronic, Abbott Vascular, Entourage Medical Technologies, and Abiomed; and lecture honoraria from Eli Lilly. Dr. Babaliaros has received consulting fees from DirectFlow Medical, Symetis, and St. Jude Medical. Dr. Williams has received consulting fees from Edwards Lifesciences. Dr. Leon is a nonpaid member of the Scientific Advisory Board of Edwards Lifesciences; and has received travel reimbursement from Edwards for activities related to his participation on the Executive Committee of the PARTNER trial. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- CI
- confidence interval
- HR
- hazard ratio
- TAVR
- transcatheter aortic valve replacement
- TF
- transfemoral
- VC
- vascular complication(s)
- Received May 14, 2012.
- Revision received June 29, 2012.
- Accepted July 2, 2012.
- American College of Cardiology Foundation
