Computational modeling identifies embolic stroke of undetermined source patients with potential arrhythmic substrate

  1. Savannah F Bifulco
  2. Griffin D Scott
  3. Sakher Sarairah
  4. Zeinab Birjandian
  5. Caroline H Roney
  6. Steven A Niederer
  7. Christian Mahnkopf
  8. Peter Kuhnlein
  9. Marcel Mitlacher
  10. David Tirschwell
  11. WT Longstreth
  12. Nazem Akoum  Is a corresponding author
  13. Patrick M Boyle  Is a corresponding author
  1. Department of Bioengineering, University of Washington, United States
  2. Division of Cardiology, University of Washington, United States
  3. Department of Neurology, University of Washington, United States
  4. School of Biomedical Engineering and Imaging Sciences, King’s College London, United Kingdom
  5. Department of Cardiology, Klinikum Coburg, Germany
  6. Department of Epidemiology, University of Washington, United States
  7. Center for Cardiovascular Biology, University of Washington, United States
  8. Institute for Stem Cell and Regenerative Medicine, University of Washington, United States
6 figures, 1 table and 1 additional file

Figures

Figure 1 with 1 supplement
Model generation.

(A) Reconstruction of LA geometry with anatomical features labeled (RIPV/RSPV/LIPV/LSPV, right/left inferior/superior pulmonary veins; LAA, LA appendage). The LA is modeled as a bilayer comprising …

Figure 1—figure supplement 1
LA subdivision scheme.

(A) Alpha coordinate from universal atrial coordinate (UAC) system mapped on to a representative LA model. (B) Beta coordinate from UAC mapped on to a representative LA model. (C) 2D representation …

Summary of patient-derived model fibrosis with respect to RD inducibility.

(A) Histogram of AFib (22/45) and ESUS (24/45) inducible patients. Inducibility was not significantly different by χ2 test. (B) Patients with ESUS and AFib arranged by percentage of LA fibrosis. …

Figure 2—source data 1

Spreadsheet including source data underlying Figure 2.

ESUS and AFib patient ID numbers; inducibility status (i.e., whether RD-sustained arrhythmia was observed in the corresponding patient-specific LA model); and LA volumetric fibrosis burden (%) extracted directly from clinical report.

https://cdn.elifesciences.org/articles/64213/elife-64213-fig2-data1-v2.xlsx
Summary of RD characteristics between ESUS and AFib models.

(A) Boxplot of fibrosis percentage in ESUS and AFib models where reentry was induced (ESUS: N = 24, IQR = 10.6; AFib: N = 22, IQR = 5) and where reentry was not induced (ESUS: N = 21, IQR = 5; AFib …

Figure 3—source data 1

Spreadsheet including source data underlying Figure 3.

For each LA model used in the study: number of pacing sites that induced RD-sustained arrhythmia; number of unique RD locations; percentage of tissue significantly depolarized in quiescent state, as presented in Figure 3F.

https://cdn.elifesciences.org/articles/64213/elife-64213-fig3-data1-v2.xlsx
Maps of fibrotic tissue distribution (left) and activation time (right) for ESUS and AFib models in which pacing succeeded (rows 1–2) or failed (row 3) to induce RD-driven arrhythmia.

Black arrows indicate directions of wavefront propagation in RDs. Double lines indicate sites of conduction block. Black-shaded regions in activation maps indicate locations where activation did not …

Summary of IdS and RD localization characteristics.

(A) Region-wise IdS for both ESUS and AFib LA models. (B) Heat map of the regions in which triggers are most likely to induce arrhythmias depicted as representative ESUS and AFib models. (C) …

Figure 5—source data 1

Spreadsheet including source data underlying Figure 5.

For each episode of RD-sustained arrhythmia induced in each patient-specific LA model, the location of the pacing associated pacing site and the LA region in which the RD ultimately anchored are provided.

https://cdn.elifesciences.org/articles/64213/elife-64213-fig5-data1-v2.xlsx
Summary of the region-by-region extent of tissue with a spatial fibrosis pattern (as characterized by local density and entropy) associated with RD localization (i.e., pro-RD tissue).

(A) Maps of pro-RD tissue for representative AFib and ESUS cases, including boundaries between regions. (B) Region-wise extent of pro-RD tissue in inducible AFib and ESUS models, depicted as …

Figure 6—source data 1

Spreadsheet including source data underlying Figure 6.

For each LA model used in the study, proportion of pro-RD tissue (as calculated by previously established machine learning-based analysis of fibrosis spatial pattern) in each LA region.

https://cdn.elifesciences.org/articles/64213/elife-64213-fig6-data1-v2.xlsx

Tables

Table 1
Patient characteristics in ESUS and AFib groups.
ESUS (N = 45)AFib (N = 45)p value
Age, years60 ± 1662 ± 120.504
Female, %44.0%32.8%0.275
BMI, kg/m227.6 ± 4.329.5 ± 5.90.08
CHA2DVASc score2.01.90.345
CHF, n14.3%18.4%0.599
Hypertension, n68.5%61.2%0.468
Diabetes mellitus, n20.4%12.2%0.292
CAD, n18.4%18.4%1.000
Smoking, n32%28%0.679
LA fibrosis, %13.6 ± 6.2%14.2 ± 4.5%0.91
LA surface area, cm2109 ± 26134 ± 400.0007
LA volume index, mL/m260 ± 2957 ± 260.607

Additional files

Download links