D-PO02-029 - Patient-Specific Computational Modeling Reveals Pre-Clinical Arrhythmic Substrate For Atrial Fibrillation In Cryptogenic Stroke Patients (ID 988)
Background: Late-gadolinium enhanced (LGE)-MRI has revealed atrial fibrotic remodeling in patients with embolic stroke of unknown source (ESUS). Curiously, ESUS patient fibrosis levels are the same as those observed in AF. A potential explanation is that ESUS patients have a fibrotic pre-clinical arrhythmic substrate, but lack the triggers needed to initiate AF.
Objective: Use computational simulations conducted in patient-specific models to determine if fibrotic remodeling in ESUS has the capacity to sustain reentrant drivers.
Methods: ESUS (per standard criteria) was verified by a neurologist. 45 patients had LGE-MRI for fibrosis assessment within 3 months of stroke. Fibrotic left atrial models were built from LGE-MRI scans. Fiber orientations were mapped onto epi/endo surfaces using universal atrial coordinates. Burst pacing from 15 sites was used to test inducibility of reentry (Fig A).
Results: In 22/45 (49%) ESUS models, we observed episodes of sustained reentry that were similar to those observed in previous AF models. Fibrosis burden was significantly higher for patients in whom simulations suggested the existence of pre-clinical arrhythmic substrate (Fig B). However, there were several examples of models that defied inducibility expectations (e.g., reentry observed in low-fibrosis models; Fig C).
Conclusion: In this modeling study, pro-arrhythmic properties of fibrosis in ESUS and AF (as seen in our prior work) are indistinguishable. This suggests some ESUS patients have pre-clinical substrate but do not have AF due to a lack of suitable triggers. Ongoing clinical follow-up will be used to test whether abundance of pre-clinical substrate is predictive of incident AF.