Abstract
Background
Limited studies are available which aim to identify patient populations that would potentially benefit from the use of cerebral embolic protection devices (CPDs) during transcatheter aortic valve replacement (TAVR). We aimed to analyze the impact of CPD use during TAVR among patients with atrial fibrillation (AF).
Methods
Data on adult TAVR patients with a concomitant diagnosis of AF was obtained from the 2017-2020 National Readmissions Database. Stroke, major stroke, in-hospital mortality, and 30-day readmission rates were compared between the CPD and no-CPD cohorts in a propensity score matched analysis. Association of CPD use with adverse events was analyzed using multivariable logistic regression models.
Results
Of 100,928 eligible TAVR patients with AF, CPD was used in 6.9% of patients with a mean age of 80 years. CPD use was independently associated with lower overall stroke (1.7% vs. 2.2%; odds ratio [OR] 0.81 [95% CI 0.68-0.98]; p = 0.032), major stroke (1.2% vs. 1.8%; OR 0.69 [0.55-0.86]; p = 0.001), in-hospital mortality (0.9 vs. 1.5%; OR 0.56 [0.43-0.72]; p < 0.001), and lower 30-day readmission rates (12.7% vs. 14.7%; OR 0.87 [0.81-0.94]; p < 0.001). Reduction in adverse events with CPD was noted in high-volume but not in low-volume TAVR centers.
Conclusions
The present point towards clear benefits of CPD use among patients with AF undergoing TAVR. In anatomically eligible patients, the potential benefit of debris capture may be considered especially as younger and lower risk patients become eligible for TAVR. Data from future trials and registries are required to further corroborate our findings.
Highlights
- •
Stroke remains a formidable cause of morbidity, mortality, and resource utilization in transcatheter aortic valve replacement (TAVR) patients.
- •
Sentinel cerebral embolic protection device (CPD) is designed to capture and remove debris thereby reducing the risk of stroke.
- •
There is an interest in identifying patients who may benefit from the use of the CPD.
- •
Patients with atrial fibrillation undergoing TAVR with CPD had lower overall stroke, and major stroke.
- •
Thirty-day all cause readmissions were lower in patients with atrial fibrillation undergoing TAVR with CPD use.
- •
Future studies from other registries and clinical trials are required to corroborate our findings and to identify candidates for CPD use.
Introduction
Stroke remains a devastating and feared complication of the revolutionary transcatheter aortic valve replacement (TAVR) procedure, affecting nearly 2% of TAVR patients. , Post-TAVR stroke is associated with increased mortality. , In order to decrease the potential complication of stroke relating to the TAVR procedure, cerebral embolic protection devices (CPDs) have been developed.
The SENTINEL CPD (Boston Scientific Corp, Boston, Massachusetts) was approved by the United States Food and Drug Administration for stroke prevention during TAVR following results from the SENTINEL randomized controlled trial (RCT) which demonstrated the safety of the device in 363 high-risk patients with aortic stenosis (AS). , The recently published results of the PROTECTED TAVR RCT (Stroke PROTECTion With SEntinel During Transcatheter Aortic Valve Replacement) further demonstrated the efficacy of the CPD device in reducing disabling stroke among 3000 TAVR patients across 50 global sites, at all surgical levels (0.5% in protected vs. 1.3% in unprotected, p = 0.02). However, the RCT did not aim to identify patients with specific underlying comorbidities who would benefit from CPD use.
Patients with atrial fibrillation (AF) are known to be at a high-risk for stroke, possibly due to cardioembolic events in the left atrial appendage. In an effort to identify candidates who may potentially benefit from CPD, we aimed to assess the outcomes of patients with AF who underwent TAVR with and without the use of CPD.
Methods
Data Source
The present study was conducted from data obtained from Nationwide Readmissions Database (NRD) (2017-2020) maintained by the Healthcare Cost and Utilization Project (HCUP) of the Agency for Healthcare Research and Quality in the United States. The data set is the largest publicly available database representing ∼52% of all US hospitalizations. Patients are tracked using verifiable linkage numbers during a calendar year, i.e., from January to the end of December. Patient characteristics, diagnoses, and procedures are coded in the database using the International Classification of Diseases (ICD) ninth and 10th Revision Clinical Modification and Procedure Coding System codes. Since the data are publicly available and contain deidentified information, approval from the institutional review board was not required for this study. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist was used when writing the present report.
Study Design
Supplementary Table 1 lists the ICD-10 Clinical Modification and Procedure Coding System codes used for the present study. Data on adult transfemoral TAVR recipients were extracted using verified ICD-10 codes (‘02RF37Z,’ ‘02RF38Z,’ ‘02RF3JZ,’ and ‘02RF3KZ’). Patients with concomitant surgery (surgical aortic valve replacement, coronary bypass surgery, or thoracic aorta surgery) were excluded, and only those with a diagnosis of AF were included in the analysis. Hospitals were categorized into metropolitan teaching hospitals vs others.
Definitions and Outcomes
The variable DISPUNIFORM provided by the HCUP contains information on the discharge disposition of patients and is coded from ‘1 to 7’ and ’20,’ corresponding to routine home/self-care discharge, ‘20’ representing death, and the rest representing discharge to another hospital/nursing facility (short-term hospital, skilled nursing facility, intermediate care facility, another type of facility, or home health care). This information was used to define “major stroke” in the present study as stroke which resulted in death or discharge to another hospital/nursing facility.
The variable HOSP_NRD provided by the HCUP was used for the calculation of annual hospital procedural volumes. Hospitals were categorized into low-volume TAVR centers if the number of procedures performed was ≤99 TAVR procedures in each year, and high-volume TAVR centers if the number of procedures performed was ≥100 TAVR procedures.
Overall stroke, major stroke, length of stay, routine discharges, in-hospital mortality, and 30-day readmission outcomes were compared between patients who underwent TAVR with and without the use of a CPD (‘X2A5312’). The weight variable provided by the HCUP, DISCWT , was used in order to obtain national estimates.
Statistical Analysis
We compared patient characteristics and outcomes between the CPD and no-CPD cohorts using Fisher exact text or chi-square test for categorical variables, and Mann-Whitney U test or Student t test continuous variables. To mitigate the risk of confounding and selection bias, a propensity score matching model was developed using logistic regression to derive 2 matched groups. The MatchIt program in R statistical software was used to perform a 1:2 nearest neighbor matching, using a caliper of 0.1. Details of the propensity matching appear in the Supplementary Materials ( Figures S1 and S2 ). All variables were also included in a multivariable logistic regression model in order to analyze the impact of CPD on overall stroke, major stroke, in-hospital mortality, and 30-day readmission rates among patients with AF undergoing TAVR. In the present study, a 2-sided p value of <0.05 was considered statistically significant. Data extraction and statistical analyses were performed using SAS (SAS Institute Inc, Cary, North Carolina), SPSS Statistics version 29 (IBM SPSS, Armonk, New York), and RStudio.
Results
Study Cohort
Figure 1 depicts the patient selection for the present study. Of 271,804 eligible transfemoral TAVR patients, 100,928 (37.1%) patients had an underlying diagnosis of AF and were included in the final analysis. The overall mean (SD) age of patients was 80.4 (7.7) years, with the majority being males (58.5%), being treated at metropolitan teaching hospitals (88.2%) ( Table 1 ). CPD was used in 6916 (6.9%) of patients with a mean age of 80 years. CPD vs no-CPD patients were more commonly males (62.6% vs. 58.2%), with a lesser burden of comorbidities such as diabetes with chronic complications (14.3% vs. 17.4%), chronic pulmonary disease (24.9% vs. 26.3%), carotid artery disease (5.6% vs. 6.3%), end-stage renal disease on dialysis (2.9% vs. 4.0%), and prior coronary bypass surgery (15.6% vs. 17.7%).
| Characteristics | Unmatched | PS-matched | |||||
|---|---|---|---|---|---|---|---|
| Overall N = 100,928 | No CPD N = 94,012 (93.1%) | CPD N = 6916 (6.9%) | p value | No CPD N = 15,826 | CPD N = 6916 | p value | |
| Mean age in y (SD) | 80.42 (7.69) | 80.44 (7.69) | 80.19 (7.79) | 0.255 | 80.36 (7.63) | 80.19 (7.79) | 0.418 |
| Females | 41,873 (41.5) | 39,289 (41.8) | 2584 (37.4) | <0.001 | 5974 (37.7) | 2584 (37.4) | 0.581 |
| Bicuspid aortic valve | 1190 (1.2) | 1032 (1.1) | 158 (2.3) | <0.001 | 358 (2.3) | 158 (2.3) | 0.918 |
| Prior valve surgery | 5278 (5.2) | 4958 (5.3) | 320 (4.6) | 0.020 | 654 (4.1) | 320 (4.6) | 0.090 |
| Metropolitan teaching hospital | 88,971 (88.2) | 82,467 (87.7) | 6504 (94.0) | <0.001 | 14,933 (34.4) | 6504 (94.0) | 0.348 |
| Hypertension | 92,376 (91.5) | 86,038 (91.5) | 6338 (91.6) | 0.720 | 14,434 (91.2) | 6338 (91.6) | 0.280 |
| Diabetes without chronic complications | 12,789 (12.7) | 11,868 (12.6) | 921 (13.3) | 0.095 | 2141 (13.5) | 921 (13.3) | 0.665 |
| Diabetes with chronic complications | 17,347 (17.2) | 16,359 (17.4) | 988 (14.3) | <0.001 | 2269 (14.3) | 988 (14.3) | 0.914 |
| Chronic pulmonary disease | 26,450 (26.2) | 24,731 (26.3) | 1719 (24.9) | 0.008 | 3782 (23.9) | 1719 (24.9) | 0.123 |
| Carotid artery disease | 6344 (6.3) | 5959 (6.3) | 385 (5.6) | 0.011 | 783 (4.9) | 385 (5.6) | 0.052 |
| ESRD on dialysis | 3921 (3.9) | 3721 (4.0) | 200 (2.9) | <0.001 | 439 (2.8) | 200 (2.9) | 0.622 |
| Prior CABG | 17,699 (17.5) | 16,620 (17.7) | 1079 (15.6) | <0.001 | 2354 (14.9) | 1079 (15.6) | 0.159 |
| Prior PCI | 21,338 (21.1) | 19,825 (21.1) | 1513 (21.9) | 0.212 | 3438 (21.7) | 1513 (21.9) | 0.799 |
| Prior PPM | 15,422 (15.3) | 14,341 (15.3) | 1081 (15.6) | 0.402 | 2322 (14.7) | 1081 (15.6) | 0.062 |
| Prior CVA | 600 (0.6) | 560 (0.6) | 40 (0.6) | 0.857 | 70 (0.4) | 40 (0.6) | 0.174 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
