A novel approach to mapping the atrial ganglionated plexus network by generating a distribution probability atlas

Min-young Kim, Markus B. Sikkel, Ross J. Hunter, Guy A. Haywood, David R. Tomlinson, Muzahir H. Tayebjee, Rheeda L. Ali, Chris D. Cantwell, Hanney Gonna, Belinda Sandler, Elaine Lim, Guy Furniss, Dimitrios Panagopoulos, Gordon A. Begg, Gurpreet Dhillon, Nicola J. Hill, James O’Neill, Dárrel P. Francis, Phang Boon Lim, Nicholas S. Peters, Nick Linton, Prapa KanagaratnamJournal of Cardiovascular Electrophysiology 2018 · DOI

TL;DR

This paper introduces a novel method to map the atrial ganglionated plexus (GP) network in the human heart by creating a probability atlas. This atlas, generated from high-frequency stimulation (HFS) data across multiple patients, identifies key regions where these nerve clusters are most likely to be found, offering a more targeted approach to understanding and potentially treating atrial fibrillation (AF).

What Problem Does This Paper Address?

Atrial fibrillation (AF) is a common heart rhythm disorder, and the autonomic nervous system (ANS), particularly the intrinsic cardiac ANS via ganglionated plexuses (GPs), is increasingly implicated in its pathogenesis. GPs are clusters of nerve cells located on the epicardial surface of the heart that influence heart rate and rhythm. While it's known that stimulating these GPs can cause atrioventricular (AV) dissociation (a slowing or stopping of the heart rate), their precise anatomical distribution and variability across individuals remain challenging to map comprehensively. Current methods often rely on post-mortem anatomical studies or limited functional mapping, leaving a gap in understanding the functional autonomic network in living patients, especially in relation to AF ablation procedures. This paper aims to address this by developing a method to create a functional probability atlas of these AVD-GPs in the left atrium.

Methods

The researchers employed a novel approach combining electrophysiological mapping with advanced data processing. They used a high-frequency stimulation (HFS) technique with a specialized catheter during AF ablation procedures in 28 patients. This HFS was applied to various points on the endocardial surface of the left atrium. The key innovation lies in how they analyzed the responses:

  • AVD-GP Identification: They defined "atrioventricular dissociating GPs" (AVD-GPs) as sites where HFS caused either ventricular asystole (complete stop) or significant bradycardia (slowing of heart rate) that was significantly beyond normal variability. They further classified these into "asystole AVD-GPs" (A-AVD-GPs) and "bradycardia AVD-GPs" (B-AVD-GPs) based on the severity of the response.
  • 3D Mapping and Registration: The 3D locations of all tested HFS sites were recorded using the CARTO electroanatomic mapping system. These individual patient maps were then coregistered and transformed onto a common reference left atrial geometry. This process allowed for the overlaying and aggregation of data from multiple patients.
  • Probability Atlas Generation: By combining the AVD-GP locations from all patients and weighting them by the uncertainty of the measurement (due to catheter movement), they generated a probability distribution atlas. This atlas visually represents the likelihood of finding AVD-GPs across the entire left atrium.

Datasets

The study utilized data from 28 patients undergoing AF ablation procedures. No new benchmark datasets were introduced; instead, the researchers collected and analyzed functional mapping data from these patients.

Key Insights and Ideas

  • Functional Mapping is Feasible: The study demonstrates that it's possible to functionally map the entire left atrium for AVD-GPs using endocardial HFS before AF ablation, providing a practical clinical application.
  • GP Distribution is Highly Variable: A significant finding is the substantial variability in the number and precise location of AVD-GPs between individual patients. This highlights the need for patient-specific mapping rather than relying solely on generalized anatomical descriptions.
  • Key Regions Identified: Despite variability, the aggregated data revealed three distinct regions in the posterior wall of the left atrium with a significantly higher probability (>20%) of containing AVD-GPs. These regions align with previous histological findings, validating the functional mapping approach.
  • Two Types of AVD-GP Responses: The observation of both asystole and bradycardia responses suggests that while the underlying GP network might be interconnected, the functional output can differ, possibly due to proximity to the AV node or the specific neural architecture activated.

Key Results

  • A total of 2108 HFS locations were tested across 28 patients.
  • 283 sites (13%) demonstrated AVD-GP effects.
  • On average, patients had 10 AVD-GPs (interquartile range 11.5).
  • 80% of AVD-GPs produced asystole (A-AVD-GPs), and 20% produced bradycardia (B-AVD-GPs).
  • The probability atlas identified three high-probability regions for AVD-GPs: the inferoseptal portion of the posterior wall (41%), the right superior pulmonary vein antrum (31%), and the base of the right inferior pulmonary vein in the posterior wall (30%).
  • The average registration error for mapping data onto the reference shell was 2.7 ± 1.0 mm.

Limitations and Future Work

  • General Anesthesia: All patients received general anesthesia, which may have influenced the threshold of GP response to HFS. Future studies could explore mapping under conscious sedation.
  • Left Atrium Focus: The study focused solely on the left atrium. The right atrium also contains GPs, and a comprehensive understanding would require mapping both atria.
  • Autonomic Modification: While the study identifies AVD-GP locations, it doesn't directly test the efficacy of ablating these sites for AF treatment. Future work could involve targeted ablation based on this atlas and evaluating its impact on AF recurrence.
  • Interpretation of AVD-GP Types: The paper suggests that the difference between A-AVD-GPs and B-AVD-GPs might be due to distance from the atrial lumen rather than fundamentally different GP types. Further research could clarify this.

The authors suggest that this atlas can serve as a guide for patient-specific mapping and could become an important step in autonomic modification procedures for AF.

Summary updated

Abstract

It is feasible to map the entire left atrium for AVD-GPs before AF ablation. Aggregated data from multiple patients, producing a distribution probability atlas of AVD-GPs, identified three regions with a higher likelihood for finding AVD-GPs and these matched the histological descriptions. This approach could be used to better characterize the autonomic network.
BibTeX
@article{kim2018novel,
  author = {Min-young Kim and Markus B. Sikkel and Ross J. Hunter and Guy A. Haywood and David R. Tomlinson and Muzahir H. Tayebjee and Rheeda L. Ali and Chris D. Cantwell and Hanney Gonna and Belinda Sandler and Elaine Lim and Guy Furniss and Dimitrios Panagopoulos and Gordon A. Begg and Gurpreet Dhillon and Nicola J. Hill and James O’Neill and Dárrel P. Francis and Phang Boon Lim and Nicholas S. Peters and Nick Linton and Prapa Kanagaratnam},
  title = {{A novel approach to mapping the atrial ganglionated plexus network by generating a distribution probability atlas}},
  journal = {Journal of Cardiovascular Electrophysiology},
  year = {2018},
  doi = {10.1111/jce.13723},
}
Figure