Mechanistic Relationships Between Fibrosis, Fibrillation, and Stroke: Multi-Scale, Multi-Physics Simulations

The main goals of this project are to identify mechanisms underlying thrombogenesis in patients with left atrial (LA) fibrosis
and to validate this new knowledge via a prospective proof-of-concept clinical study. Atrial fibrillation (AFib) affects millions
of Americans and carries a five-fold increased risk of stroke, a leading cause of mortality and morbidity. Around 30% of all
ischemic strokes are caused by thromboembolism in AFib patients. In patients without AFib, embolic strokes of undetermined
source (ESUS) account for an additional 30% of ischemic strokes. Current stroke risk stratification tools in AFib and ESUS (e.g.,
CHA2DS2-VASc) are deficient in predictive accuracy, leaving many patients either under-treated for stroke prevention or over-
treated and subjected to unnecessary bleeding risk. The growing evidence that LA fibrosis serves as a mechanistic nexus
between AFib and ESUS is a very promising advance that could open new avenues for stroke prevention. However, taking
advantage of this opportunity requires detailed knowledge of the mechanism(s) by which fibrotic atria are prone to thrombosis,
with or without AFib. Fibrosis has complex structural, electrical, and contractile effects in the LA. These phenomena may
independently or synergistically influence thrombosis risk by altering LA hemodynamics, but prior work has not systematically
assessed inter-dependencies or clarified each factor’s relative importance. This is due to difficulties associated with
experimental manipulation and acquisition of clinical measurements. Advances in computational modeling offer an
unprecedented opportunity to address this critical knowledge gap. Specifically, the stage is set to create a multi-scale, multi-
physics framework that can comprehensively simulate the pro-thrombotic potential of each unique patient-specific LA
fibrosis pattern. Our central hypothesis is that LA fibrosis is a key mechanistic factor in determining each individual’s risk of
thromboembolic stroke due to structural, electrical, and contractile factors. Our approach consists of three specific aims.
Aim 1 will develop and calibrate a computational framework that integrates electrophysiological, biomechanical, and mechano-
fluidic modeling in patient-specific LA models, paying special attention to resolving the effects of fibrosis. We will parameterize
the framework using multi-modality magnetic resonance imaging acquisitions in AFib patients with prior stroke and non-AFib,
non-stroke controls. Aim 2 will use the new computational framework to systematically characterize mechanistic connections
between LA fibrosis and thrombogenesis. We will examine how each individual’s mix of fibrosis extent/pattern, LA anatomy,
and susceptibility to emergent electromechanical phenomena combine (with or without simulated AFib) to create a
thrombogenic milieu that can be characterized by computational modeling. Aim 3 will validate the mechanistic connections
between fibrosis and risk of recurrent stroke/brain microinfarction in a proof-of-concept prospective clinical study. We will
examine a high-risk cohort of ESUS patients, but notably without a current indication for oral anticoagulation. We will test if
model-predicted thrombogenic combinations of LA shape, fibrosis pattern, deranged electromechanics, and disrupted blood
flow exist in patients who experience more adverse outcomes. Our validated multi-physics modeling framework will, for the
first time, yield new insight on fibrosis-mediated stroke mechanisms, and pave the way for new treatments for millions of
patients who are borderline candidates for anticoagulation (e.g., individuals with ESUS or AFib with intermediate risk scores).

Grant Number: 5R01HL158667-04
NIH Institute/Center: NIH
Principal Investigator: Patrick Boyle