Copyright
©The Author(s) 2023. Published by Baishideng Publishing Group Inc. All rights reserved.
Prediction of permanent pacemaker implantation after transcatheter aortic valve replacement: The role of machine learning
Pradyumna Agasthi, Hasan Ashraf, Sai Harika Pujari, Marlene Girardo, Andrew Tseng, Farouk Mookadam, Nithin Venepally, Matthew R Buras, Bishoy Abraham, Banveet K Khetarpal, Mohamed Allam, Siva K Mulpuru MD, Mackram F Eleid, Kevin L Greason, Nirat Beohar, John Sweeney, David Fortuin, David R Jr Holmes, Reza Arsanjani
Pradyumna Agasthi, Hasan Ashraf, Farouk Mookadam, Nithin Venepally, Bishoy Abraham, Banveet K Khetarpal, Mohamed Allam, John Sweeney, David Fortuin, Reza Arsanjani, Department of Cardiology, Mayo Clinic, Phoenix, AZ 85054, United States
Sai Harika Pujari, Department of Internal Medicine, The Brooklyn Hospital Center, Brooklyn, NY 11201, United States
Marlene Girardo, Department of Biostatistics, Mayo Clinic, Phoenix, AZ 85054, United States
Andrew Tseng, Siva K Mulpuru MD, Mackram F Eleid, David R Jr Holmes, Department of Cardiovascular Diseases, Mayo Clinic, Rochester, MN 55905, United States
Matthew R Buras, Department of Statistics, Mayo Clinic, Phoenix, AZ 85054, United States
Kevin L Greason, Department of Cardiovascular Surgery, Mayo Clinic, Rochester, MN 55905, United States
Nirat Beohar, Mount Sinai Medical Center, Columbia University, Miami Beach, FL 33138, United States
Author contributions: Agasthi P conceptualizated, did methodology, project administration, review, and editing; Ashraf H methodology and writing the manuscript; Pujari SH did the data curation, visualization, and writing; Girardo M did the formal analysis, methodology, review, and editing; Tseng A writing, review, and editing the manuscript; Mookadam F did the methodology, project administration, review, and editing; Venepally N did the writing and data curation; Buras MR did the methodology and writing; Allam M did the data curation, visualization, and writing; Abraham B, Khetarpal BK, MD SKM, Eleid MF, Greason KL, Beohar N, Sweeney J, and Fortuin D writing, review and editing the manuscript; Holmes DRJ and Arsanjani R did the methodology, project administration, review, and editing.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Data sharing statement: The authors confirm that the data supporting the findings of this study are available.
Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See:
https://creativecommons.org/Licenses/by-nc/4.0/ Corresponding author: Sai Harika Pujari, MBBS, Department of Internal Medicine, The Brooklyn Hospital Center, 121 Dekalb Avenue, Brooklyn, NY 11201, United States. spujari@tbh.org
Received: November 25, 2022
Peer-review started: November 25, 2022
First decision: December 13, 2022
Revised: January 4, 2023
Accepted: March 1, 2023
Article in press: March 1, 2023
Published online: March 26, 2023
Processing time: 115 Days and 10.6 Hours
ARTICLE HIGHLIGHTS
Research background
For aortic stenosis, it is a fact that transcatheter aortic valve replacement use has greatly increased relative to surgical replacement with the most common complications of the procedure including atrioventricular conduction abnormalities development and permanent pacemaker requirement (PPM). Hence, it is essential to risk stratify patients for potential need of PPM implantation post-procedure. We used artificial intelligence to predict pre-procedural risk for pacemaker placement post-transcatheter aortic valve replacement at 30 d and 1 year.
Research motivation
Previous studies that evaluated risk factors associated with permanent pacemaker requirement used data for older-generation valves and also included only a limited number of variables and hence, limiting their predictive potential. Artificial intelligence does a remarkable job of predicting variables via machine learning and the same has been used in our study.
Research objectives
To predict pre-procedural risk for permanent pacemaker post-transcatheter aortic valve replacement (TAVR) at 30 d and 1 year.
Research methods
We performed a retrospective study on patients with severe symptomatic aortic stenosis who underwent transcatheter aortic valve replacement (TAVR). Gradient boosting machine learning model has been used for predicting probabilities.
Research results
For 30-d analysis, higher brachiocephalic artery to annulus distance to patient height ratio was the highest weighted characteristic that predicted PPM placement post- TAVR. Also for 1-year analysis, higher brachiocephalic artery to annulus distance to patient height ratio was the highest weighted characteristic that predicted PPM placement post- TAVR.
Research conclusions
Brachiocephalic to annulus distance to height ratio is the highest weighted predictor of PPM implantation in the study both at 30 d and 1 year and it was not been previously described in the literature.
Research perspectives
We sought to develop and have developed a risk assessment tool to predict PPM implantation post-TAVR using machine learning.
