Early Experience With Cerebral Embolic Protection During Transcatheter Aortic Valve Replacement in the United States

Despite contemporary advances in transcatheter aortic valve replacement (TAVR), procedural strokes remain a weakness of the procedure.¹ Cerebral embolic protection devices (EPDs) emerged as a potential method to decrease the risk of post-TAVR strokes. However, data supporting the use of EPDs are conflicting. In the Cerebral Protection in TAVR trial, EPDs did not reduce the rate of post-TAVR stroke, although embolic debris was captured in 99% of filters.² However, observational data and pooled analyses have suggested a potential benefit of EPDs.^3,4 This study sought to evaluate the early post–US Food and Drug Administration approval experience with the Sentinel device (Boston Scientific), the only EPD available in the United States.

The Vizient database was queried to select patients who underwent TAVR with or without EPD between January 1, 2017, and December 31, 2018, using International Statistical Classification of Diseases and Related Health Problems, Tenth Revision, Clinical Modification (ICD-10-CM) codes (TAVR: codes 02RF3JZ, 02RF3KZ, 02RF38Z, 02RF37Z, 02RF3JZ, 02RF37H, 02RF38H, 02RF3JH, and 02RF3KH; EPD: code X2A5312). The Vizient database contains deidentified administrative, clinical, and financial inpatient information of 100% of index hospitalizations at more than 400 US academic centers and their affiliated hospitals. The West Virginia University Institutional Review Board exempted this study from review because this study used publicly available deidentified data.

Patients were classified into those who had TAVR with EPD between July 1, 2017, and December 31, 2018, and those who had TAVR without EPD between January 1, 2017, and June 30, 2017. The nonoverlapping time periods were chosen to avoid the potential for undercoding of new technology. Because EPDs were not approved prior to July 2017, this strategy ensured that no patients in the no-EPD group could have been misclassified owing to undercoding.

The primary end point was any stroke or transient ischemic attack. Secondary end points were ischemic stroke, hemorrhagic stroke, fatal stroke, disabling stroke, death, length of stay, and cost. These end points were identified using previously reported ICD-10-CM codes.⁵ Propensity score matching was performed to adjust for differences in baseline characteristics. The variables used in the propensity score matching included the published factors associated with post-TAVR strokes (Table).⁶ Outcomes were compared using χ² test for categorical variables and independent samples t test for continuous variables. All P values were from 2-sided tests and results were deemed statistically significant at P < .05.

A total of 10 985 patients were included. The proportion of hospitals using EPD increased from 8.6% (12 of 139) in quarter 3 of 2017 to 32.4% (48 of 148) in quarter 4 of 2018. The proportion of patients undergoing TAVR who received EPD increased from 2.8% (136 of 4879) in quarter 3 of 2017 to 17.3% (950 of 5479) in quarter 4 of 2018. Patients in the no-EPD group had a higher prevalence of key comorbidities but similar rates of carotid stenosis and obesity (Table). The primary end point occurred in 183 of 8253 patients (2.2%) in the no-EPD group and in 48 of 2732 patients (1.8%) in the EPD group (P = .15). There were no statistically significant differences in the primary or secondary clinical end points before or after propensity score matching (Table). Mean (SD) length of stay was longer in the no-EPD group (5.0 [7.0] vs 4.1 [4.0] days; P < .001), but mean (SD) procedural and total costs were higher in the EPD group (procedural costs: $37 070 [$15 245] vs $35 527 [$19 255]; and total costs: $52 522 [$20 427] vs $51 055 [$28 009]).

In this survey of the early experience with EPD, the use of EPD was not associated with a statistically significant reduction in post-TAVR stroke or transient ischemic attack. The following should be considered while interpreting these findings. First, the study’s end points were site reported and are subject to undercoding or overcoding. However, we used the standardized stroke codes that are used by the Centers for Medicare & Medicaid Services. We also performed an internal validation in our TAVR cohort (n = 307), which showed that the codes used correctly identified all stroke events (n = 8). Second, subclinical strokes could not be assessed in this study, although this limitation applies to all available databases, including the Transcatheter Valve Therapy registry. Third, owing to the small number of events, we are unable to adjust for certain factors that might be associated with the performance of EPDs (eg, learning curve, TAVR volume, and surgical risk). Nonetheless, this study provides important insights into the early experience with EPDs in the United States and underscores the need for additional prospective registries and randomized clinical trials to confirm the suggested benefit of EPDs during TAVR.

Accepted for Publication: November 21, 2019.

Corresponding Author: Mohamad Alkhouli, MD, Department of Cardiology, Mayo Clinic School of Medicine, 200 First St SW, Rochester, MN 55905 (alkhouli.mohamad@mayo.edu).

Published Online: February 24, 2020. doi:10.1001/jamainternmed.2019.6767

Author Contributions: Dr Alqahtani and Ms Harris had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Alkhouli, Alqahtani, Harris, Rihal.

Acquisition, analysis, or interpretation of data: Alkhouli, Alqahtani, Harris, Hohmann.

Drafting of the manuscript: Alkhouli, Alqahtani.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Alqahtani, Hohmann.

Administrative, technical, or material support: Alkhouli.

Supervision: Alkhouli, Rihal.

Conflict of Interest Disclosures: None reported.