Author + information
- Received September 10, 2015
- Accepted October 24, 2015
- Published online February 22, 2016.
- Nirat Beohar, MDa,∗ (, )
- Ajay J. Kirtane, MD, SMb,
- Eugene Blackstone, MDc,
- Ron Waksman, MDd,
- David Holmes Jr, MDe,
- Sa’ar Minha, MDd,
- Oluseun Alli, MDf,
- Rakesh M. Suri, MD, DPhilc,
- Lars G. Svensson, MD, PhDc,
- Martin Leon, MDb and
- Susheel Kodali, MDb
- aColumbia University, Division of Cardiology, Mount Sinai Medical Center, Miami Beach, Florida
- bDepartment of Medicine, Columbia University Medical Center/New York Presbyterian Hospital, New York, New York
- cDepartment of Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cleveland, Ohio
- dDivision of Cardiology, Medstar Washington Hospital Center, Washington, DC
- eDepartment of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota
- fDivision of Cardiovascular Disease, University of Alabama-Birmingham, Birmingham, Alabama
- ↵∗Reprint requests and correspondence:
Dr. Nirat Beohar, Cardiac Catheterization Laboratory, Columbia University Division of Cardiology, Mount Sinai Medical Center, 4300 Alton Road, Miami Beach, Florida 33140.
Objectives The aim of this study was to examine trends in the rates of complications and outcomes of patients undergoing transfemoral transcatheter aortic valve replacement (TF-TAVR).
Background It is unknown whether an evolution of case selection or accrual of case experience over time has resulted in a change in the rates of complications and outcomes of patients undergoing TF-TAVR.
Methods TF-TAVR patients enrolled in the PARTNER (Placement of AoRTic TraNscathetER Valve Trial) nonrandomized continued access registry (N = 1,063, enrolled March 2011 to January 2012 after completion of the randomized trial) were divided into tertiles (T1 through T3) based on enrollment date. Patient characteristics and rates of adverse events were compared over time.
Results There were no significant differences in sex, New York Heart Association functional classes III/IV, diabetes, coronary artery disease, previous revascularization, pulmonary hypertension, renal disease, or liver disease. There was an increase in mean age, but a decrease in porcelain aorta, chronic obstructive pulmonary disease (including oxygen-dependent chronic obstructive pulmonary disease), previous chest wall radiation, and a slight decrease in the median Society of Thoracic Surgeons Predicted Risk of Mortality score. There was a significant decline in the frequency of patients deemed “inoperable” (cohort B) and in need for post-dilation. Percutaneous access increased significantly. There were no differences in post-procedural stroke, major bleeding, major vascular complications, or the need for aortic valve reintervention over time. The incidence of moderate/severe paravalvular regurgitation declined significantly as did all-cause mortality at 1 and 2 years.
Conclusions A significant reduction in the incidence of moderate/severe paravalvular regurgitation as well as longer term all-cause mortality was observed over time. The cause of these reductions was likely multifactorial, including improved case selection and procedural techniques and increased site experience. (THE PARTNER TRIAL [Placement of AoRTic TraNscathetER Valve Trial]; NCT00530894)
Transcatheter aortic valve replacement (TAVR) has been shown to improve survival compared with surgical aortic valve replacement in high-risk patients (1) and in inoperable patients (2,3) with severe symptomatic aortic stenosis (AS). In recent years, there has been a significant growth in the clinical adoption of TAVR in inoperable, high risk and low-to-intermediate risk patients worldwide with consequent greater operator and heart team experience with the procedure as well as with patient selection (4–6). It is estimated that >100,000 TAVRs have been performed worldwide between 2002 and 2013 (7).
Despite its growth, TAVR as a procedure is still evolving and requires further refinement to reduce complications. It is unclear whether greater adoption and experience have been associated with changes in the rates of complications and outcomes of patients undergoing TAVR. After completion of the PARTNER (Placement of AoRTic TraNscathetER Valves) 1 randomized, controlled trial and before commercial approval of the transcatheter heart valve (SAPIEN, Edwards Lifesciences, Irvine, California), additional patients were treated in a randomized continued access trial as well as in a PARTNER nonrandomized continued access registry (NRCA) with the same inclusion and exclusion criteria as the randomized PARTNER 1 trial (8). The NRCA provides a suitable population in which to study the effects of an evolution in patient selection and changes in procedural complications on the outcomes of high-risk and inoperable patients with symptomatic aortic stenosis (AS) undergoing TAVR in a real-world clinical setting. We sought to compare the incidence of these outcomes over time within the NRCA.
From March 2011 to January 2012, a total of 2,068 patients were enrolled in the PARTNER trial NRCA, of whom 1,063 patients were treated with TAVR using a transfemoral (TF) approach (TF-TAVR). These 1,063 patients were included in the present as-treated analysis and were divided into tertiles based on the date of procedure as follows: T1 (March 24, 2009 to July 21, 2010, n = 353), T2 (July 22, 2010 to March 10, 2011, n = 355), and T3 (March 11, 2011 to January 10, 2012, n = 355) (Figure 1, Table 1).
All patients had severe native trileaflet AS documented on a screening transthoracic echocardiogram within 30 days of enrollment and were evaluated by 2 surgeons for assessment of risk with surgical aortic valve replacement. Important exclusion criteria included bicuspid aortic valve disease, ejection fraction <20%, renal failure, severe mitral regurgitation, severe aortic regurgitation, recent gastrointestinal bleeding, or a recent neurological event. Complete inclusion and exclusion criteria have been presented in previous publications (2,9).
All patients undergoing TF-TAVR received either a 23- or a 26-mm balloon-expandable Edwards SAPIEN transcatheter heart valve (Edwards Lifesciences, Irvine, California). At the time of the registry, a 29-mm valve was not available. Annular assessments to determine valve size required were site determined using transthoracic echocardiography, transesophageal echocardiography, or multislice computed tomography. All patients underwent transthoracic echocardiography before discharge and at clinical follow-up assessments, including at 1 month, 6 months, and 1 year. All echocardiograms were analyzed at an independent core laboratory with methodology described previously (10). Patient characteristics and rates of periprocedural and adverse clinical events were compared over time. Clinical events were adjudicated by an independent clinical events committee. The institutional review board at each participating site approved the study, and all patients provided written informed consent.
The frequency of all-cause mortality (30 days, 6 months, 1 year, and 2 years) and 30-day nonfatal complications including any stroke, any aortic valve repeat intervention, major bleeding, and major vascular complications were reported according to a modified version of the Academic Research Consortium-1 criteria (11). These endpoints were pre-specified and adjudicated. A composite endpoint of 30-day nonfatal major complications comprised any stroke, any aortic valve repeat intervention, major bleeding, and major vascular complications was also reported.
Echocardiogram-derived paravalvular regurgitation (PVR) (graded based on either discharge or 30-day echocardiogram) was graded as none/trace, mild, moderate, or severe using semiquantitative criteria previously described (12). Briefly, PVR after TAVR/surgical aortic valve replacement was graded in accordance with the American Society of Echocardiography recommendations for native valves (13) with 1 exception. Because of the often eccentric, irregular jet and the frequent noncylindrical “spray” of the paravalvular jet contour, the parasternal short-axis view(s) was weighted more heavily than other signals in providing an integrated assessment, as follows: none, no regurgitant color flow; trace, pinpoint jet in the aortic valve; mild, jet arc length is <10% of the annulus circumference; moderate, jet arc length is 10% to 30% of the annulus circumference; severe, jet arc length is >30% of the annulus circumference.
Continuous variables were summarized as mean ± SD or as median and quartile, as appropriate. Survival curves for time-to-event variables were constructed on the basis of all available follow-up data using Kaplan-Meier estimates, and comparisons were performed using the log-rank test. A trend test was performed to determine the trend across tertiles summarized as p values in Tables 1, 2, and 3. Univariate and multivariable analyses were performed to determine baseline clinical and echocardiographic characteristics that contributed to differences in all-cause mortality between the 3 tertiles. Cox proportional hazard regression analysis was performed to determine independent predictors of 1-year all-cause mortality. The multivariable model was built by stepwise selection, with candidate variables being selected if they were of clinical interest or satisfied the entry criterion of p < 0.10 in the univariable analysis. Variables were entered with entry/stay criteria of 0.1/0.1 in a forward stepwise fashion. The proportional hazards assumption was checked in the Cox model by creating interactions of the predictors and a function of survival time, and none were found to be significant.
A 2-sided alpha level of 0.05 was used for all statistical testing. All statistical analyses were performed using SAS version 9.4 (SAS Institute Inc., Cary, North Carolina). An independent academic biostatistics group performed all data analyses. Authors had full access to the data and take full responsibility for the accuracy of the data and the decision to publish.
Patients and baseline characteristics
Patients in T1 were younger, less likely to be 80 years of age or older, and had a higher mean Society of Thoracic Surgeons Predicted Risk of Mortality score compared with T2 or T3 patients. The frequency of diabetes mellitus, dyslipidemia, hypertension, congestive heart failure (New York Heart Association functional class III or IV), previous coronary artery disease or previous coronary artery bypass surgery, renal disease (serum creatinine ≥2.0 mg/dl), liver disease, pulmonary hypertension, or the mean Mini-Mental Status Examination score was similar among the 3 groups. However, the frequency of prior stroke or transient ischemic attack, porcelain aorta, chronic obstructive pulmonary disease (COPD) including oxygen-dependent COPD, previous chest wall radiation, and chest wall deformities was higher in T1 compared with T2 or T3 patients. In terms of echocardiographic measurements, minor but statistically significant differences in left ventricular ejection fraction was found among the 3 groups; however, there were no significant differences in stroke volume index, mean transaortic gradient, or planimetered aortic valve area. Of interest, there was a significant decline in the frequency of patients deemed “inoperable” (cohort B) who underwent TAVR from T1 to T3 (Table 1).
Device and procedural success significantly improved over the course of enrollment. There were no differences over time in the size of the valve used (23 or 26 mm) for TAVR; however, there was a significant decrease in the need for post-dilation after valve placement. There was a significant increase in the use of fully percutaneous access and a decline in the need for surgical cut down for TAVR arterial access from T1 to T3 (Table 2).
All-cause mortality and post-procedural complications
Although there was no difference in all-cause mortality at 30 days (T1, T2, T3; 5.4%, 5.1%, 2.8%; p across groups = 0.10), a significant decline across tertiles was observed at 6 months (16.7%, 13.9%, 10.2%; p = 0.01), 1 year (22.7%, 21.4%, 14.4%; p = 0.005) and 2 years (34.9%, 32.6%, 24.2%; p = 0.005) (Figure 2). At 30 days of follow up, there was no difference in any stroke (3.5%, 4.0%, 3.4%; p = 0.89), major bleeding (8.6%, 6.3%, 6.2%; p = 0.38), major vascular complications (9.7%, 6.2%, 8.2%; p = 0.24), any aortic valve reintervention (0.3%, 0.9%, 1.4%; p = 0.24), or in the composite of these nonfatal major complications (17.1%, 13.3%, 15.2%; p = 0.39) (Table 3). However, there was a significant decline in the occurrence of 30-day post-procedural moderate or severe PVR after TAVR (19.2%, 13.8%, 10.1%; p = 0.004) (Table 3, Figure 3).
After adjustment for age, sex, cohort A versus B status, baseline body mass index, Society of Thoracic Surgeons score for mortality, major arrhythmia, COPD, renal disease (creatinine >2 mg/dl), baseline ejection fraction, baseline stroke volume index, baseline mean aortic valve gradient, and liver disease, enrollment in T3 was independently associated with a 33% decline in all-cause mortality compared with T1 (hazard ratio: 0.67, 95% confidence interval: 0.46 to 0.97; p = 0.036). Enrollment in T2 compared with T1 was not independently predictive of mortality (Figure 4).
The major findings of the present study are as follows. First, during the enrollment period of the PARTNER NRCA, in TF-TAVR patients, there was a significant decline in 1- and 2-year all-cause mortality. Second, patients treated in the third tertile of enrollment had less frequent baseline comorbidities and fewer high-risk patient characteristics such as previous stroke or transient ischemic attack, porcelain aorta, COPD, previous chest wall radiation, or chest wall deformities. There was also a significant decline in the percentage of patients deemed inoperable and significantly less frequent use of surgical access methods for TAVR. Third, there was a significant decline in the frequency of PVR and the need for post-dilation after valve placement, possibly due to improved valve sizing methods.
The decline in overall mortality over time in this analysis is notable. A previous analysis of PARTNER 1 patients undergoing TF-TAVR found that the NRCA population had a significantly lower overall 1-year all-cause mortality rate (19.0% vs. 25.3%; p = 0.009) compared with randomized, controlled trial patients (8). In the present analysis of the NRCA population undergoing TF-TAVR, we found that all-cause 1-year mortality continued to decline from the first enrollment cohort (T1) to the third enrollment cohort (T3) (22.7% vs. 14.4%; p = 0.005) and that this significant decline was sustained at 2-year follow-up (34.9% vs. 24.2%; p = 0.005). To put these results into perspective, the all-cause 1-year mortality achieved in the T3 cohort (14.4%) was lower than the corresponding 1-year all-cause mortality (as-treated analysis) of the PARTNER 1A TF cohort (21.4%) (2) and CoreValve (Medtronic, Minneapolis, Minnesota) extreme-risk cohort (24.3%) (14) and was similar to that of the CoreValve high-risk cohort (14.2%) (1). In the present analysis, mean Society of Thoracic Surgeons score was higher in the T3 cohort (10.7 ± 2.9) compared with the CoreValve high-risk cohort (7.3 ± 3.0) (1) and similar to the CoreValve extreme-risk cohort (10.3 ± 5.5) (14) and lower than the PARTNER 1A TF cohort (11.7 ± 3.3) (2). Additionally, the 2-year all-cause mortality (as-treated analysis) in the inoperable (43.3%) (15) and high-risk TF (30.7%) (16) cohorts of the PARTNER 1 trial was higher than that in the T3 cohort.
The significant decline in all-cause 1- and 2-year morality found in the present analysis was likely multifactorial, with causes including the enrollment of patients with more favorable clinical risk profiles over the course of the NRCA enrollment period. We found that significant changes occurred in several important clinical characteristics of patients during the enrollment period. Baseline comorbidities were reduced, and fewer patients with high-risk characteristics such as previous stroke or transient ischemic attack, porcelain aorta, COPD, oxygen-dependent COPD, previous chest wall radiation, and chest wall deformities, were enrolled. Of these differences, COPD and previous stroke are especially notable. The incidence of chronic lung disease in large TAVR registries is 21% to 43% (2,9,17–21), and it has previously been shown that patients with chronic lung disease, especially oxygen dependent and those with pulmonary hypertension, experience higher mortality after TAVR compared with those without (8). Previous stroke has also been identified as an important risk factor for increased mortality after TAVR (18).
We found a significant and steady decline in the frequency of patients considered being inoperable over the course of enrollment. Cohort A (high-risk) patients in the PARTNER 1 randomized trial had a less adverse profile of baseline comorbidities compared with cohort B (inoperable) patients. Underscoring the important effect of baseline comorbidities on subsequent outcomes of patients undergoing TAVR, we found classification as a cohort A (high-risk) patient compared with cohort B (inoperable) to be associated with reduced risk of 1-year mortality after TF-TAVR (2,9). It is possible that over the course of the PARTNER trial, as heart team and surgical experience with this complex patient population accrued, the types of patients initially considered to be inoperable increasingly became labeled as high risk. However, the decline in important comorbidities between cohorts refutes this theory, instead suggesting that subsequent enrollment cohorts were indeed high risk rather than “reclassified” inoperable patients.
Despite limitations with calibration and discrimination in predicting mortality of patients undergoing TAVR (22), a markedly increased Society of Thoracic Surgeons Predicted Risk of Mortality score, reflective of more adverse demographic and clinical features, has previously been shown to predict increased mortality or lack of benefit with TAVR (15,23).
We found a significant decline in moderate or severe PVR after TF-TAVR and also a reduction in post-dilation after valve deployment. The detrimental effect of PVR has been reported in studies of both the SAPIEN (Edwards Lifesciences) and CoreValve (Medtronic) devices (18,21,24–27), and it seems likely that the decrease in all-cause mortality over time may be related to the reduction of this important complication. The decline in PVR observed in the present analysis was likely related to better valve sizing and positioning, as suggested by the decreasing need for post-dilation from T1 to T3. Advancements in imaging techniques for better sizing and placement of the TAVR prosthesis, development of newer TAVR devices to minimize PVR, and percutaneous techniques to address residual PVR are likely to minimize this complication and its adverse effects and to further lower post-TAVR mortality (27–30).
A previous analysis showed that periprocedural nonfatal complications (any stroke, any aortic valve repeat intervention, major bleeding, and major vascular complications) were independently predictive of all-cause 1-year mortality (23); however, we found no significant decrease in the frequency of these in the 3 cohorts. Fully percutaneous access rather than surgical cut down significantly increased in the 3 cohorts (30.9% in T1 to 62.1% in T3). It is likely that further increases in the use of fully percutaneous techniques with lower profile devices, better dedicated closure devices, and increasing operator experience may continue to lower the occurrence of major vascular complications and associated major bleeding, both of which are known to increase mortality after TAVR (31,32). As TAVR prostheses have decreased in profile and experience with percutaneous access has increased, TF access (in most cases, using a fully percutaneous approach) has increased. In the recently published CoreValve trials, a TF approach was used in 82.8% of the high-risk population (as-treated population) and 99.3% of the extreme-risk population (1,14); this is also consistent with the U.K. TAVI registry (21). Periprocedural complications not only increase mortality but also the cost of TAVR, accounting for 25% of nonimplant-related hospital costs in 1 analysis (33). Avoidance of complications remains an important goal to decrease mortality and make TAVR cost-effective in inoperable and high-risk patients. There is clearly room for improvement, and procedural complications can likely be decreased with improvement in device profiles, better sizing techniques, decreasing frequency of strokes with devices and optimal antithrombotic/antiplatelet therapies, fully percutaneous access, and better closure devices (34).
Reducing late (1- and 2-year) mortality after TF-TAVR is highly dependent on strategic case selection, the enrollment of patients with fewer high-risk characteristics, and avoidance of “futile” patients, lowering the frequency of moderate/severe periprocedural PVR and reducing nonfatal major procedural complications.
The impact of increased operator and institutional experience and evolving technical or procedural enhancements such as improved delivery systems and percutaneous access or closure techniques on all-cause mortality after TAVR were not assessed in this analysis. The present analysis is a post hoc analysis from a nonrandomized registry and has inherent limitations related to such an approach.
All-cause mortality after TAVR can be lowered. In the present analysis, selection of patients with fewer adverse clinical features (likely reflecting greater heart team experience and corresponding better clinical judgment in selecting patients for TAVR) and lowering of PVR (possibly due to improved valve sizing methods) were related to a significant decline in post-TAVR mortality through 2 years of follow-up. We found no reduction in the frequency of nonfatal major complications, however, which may be a target for further reduction of late mortality after TAVR.
WHAT IS KNOWN? Although TAVR has been shown to improve survival in high-risk and inoperable patients with severe symptomatic AS, all-cause mortality after successful TAVR remains high. It is not known whether an evolution of case selection or accruing case experience over time has resulted in a change in the rates of complications and outcomes of patients undergoing TF-TAVR.
WHAT IS NEW? We found that a significant reduction in the incidence of longer term all-cause mortality as well as paravalvular leakage over time among patients undergoing TF-TAVR during the enrollment period in the PARTNER NRCA. The cause of these reductions is likely multifactorial, including improved case selection and valve sizing techniques.
WHAT IS NEXT? We found no reduction in the frequency of nonfatal major complications, however, and techniques to reduce their frequency may further reduce late mortality after TAVR.
The authors thank Girma Minalu Ayele and Feifan Zhang of the Cardiovascular Research Foundation for statistical support in the preparation of this manuscript.
The PARTNER trial was funded by Edwards Lifesciences. Dr. Suri is a national PI for the PERCEVAL trial (Sorin Medical), on the Steering Committee for the Portico Trial (St. Jude Medical), and co-investigator for the PARTNER II (Edwards Lifesciences) and COAPT (Abbott) trials. Dr. Svensson holds equity in Cardiosolutions and ValvXchange; has intellectual property rights/royalties from Posthorax; and is an unpaid member of the PARTNER Trial Executive Committee. Dr. Leon is an unpaid member of the PARTNER Trial Executive Committee. Dr. Kodali is a consultant for Edwards Lifesciences; and serves on the Scientific Advisory Board of Thubrikar Aortic Valve. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- aortic stenosis
- chronic obstructive pulmonary disease
- nonrandomized continued access registry
- paravalvular regurgitation
- transcatheter aortic valve replacement
- transfemoral transcatheter aortic valve replacement
- Received September 10, 2015.
- Accepted October 24, 2015.
- 2016 American College of Cardiology Foundation
- Mylotte D.,
- Osnabrugge R.L.,
- Windecker S.,
- et al.
- Piazza N.,
- Kalesan B.,
- van Mieghem N.,
- et al.
- Fearon W.F.,
- Kodali S.,
- Doshi D.,
- et al.
- Leon M.B.,
- Piazza N.,
- Nikolsky E.,
- et al.
- Hahn R.T.,
- Pibarot P.,
- Stewart W.J.,
- et al.
- Popma J.J.,
- Adams D.H.,
- Reardon M.J.,
- et al.
- Thomas M.,
- Schymik G.,
- Walther T.,
- et al.
- Tamburino C.,
- Capodanno D.,
- Ramondo A.,
- et al.
- Moat N.E.,
- Ludman P.,
- de Belder M.A.,
- et al.
- Beohar N.,
- Whisenant B.,
- Kirtane A.J.,
- et al.
- Beohar N.,
- Zajarias A.,
- Thourani V.H.,
- et al.
- Dvir D.,
- Waksman R.,
- Barbash I.M.,
- et al.
- Toggweiler S.,
- Humphries K.H.,
- Lee M.,
- et al.
- Lerakis S.,
- Hayek S.S.,
- Douglas P.S.
- Genereux P.,
- Head S.J.,
- Hahn R.,
- et al.
- Genereux P.,
- Webb J.G.,
- Svensson L.G.,
- et al.
- Genereux P.,
- Cohen D.J.,
- Williams M.R.,
- et al.
- Arnold S.V.,
- Lei Y.,
- Reynolds M.R.,
- et al.