Accelerated Intrinsic Beating Rate in Heterogeneously Coupled Stem Cells May Underlie Focal Ventricular Tachycardia in Regenerative Cardiac Therapy

Introduction: Ischemic heart disease remains a significant global health burden. Following a myocardial infarction (MI), over one billion cardiomyocytes are lost and are replaced by scar tissue. Cardiac regeneration using human pluripotent stem cell-derived cardiomyocyte (hPSC-CM) transplantation into infracted areas is a promising therapy but can lead to focal ventricular tachycardia (VT). Mechanistic understanding of VT following cell injection is hindered by the lack of in-silico models that recapitulate phenomena seen in vivo. Current cell-scale hPSC-CM models beat between 0.58 Hz and 1.9 Hz. We hypothesize that the spontaneous beating of the in vitro data used to calibrate these hPSC-CM is too slow to sustain focal VT in vivo. Here, we build upon our prior computational framework to assess the minimum rate of spontaneous hPSC-CM excitation needed for grafts to overwhelm sinus rhythm and sustain focal VT.

Materials and

Methods: A human ventricular slice model with border zone, healthy myocardium, and infarct was generated with six superimposed histology-derived hPSC-CM injection patterns (Fig A). Four cell-scale hPSC-CM models ranging from 1.1 Hz to 4 Hz were tested (Fig B). Ventricular, border zone, and hPSC-CM cellular models were used in the associated regions; infarcted tissue was modeled as non-conductive. An established graft-host coupling model was used to probe the likelihood of ectopic focal VT in a probabilistic manner. A population of 40 variants were generated with 7.5% connected graft-boundary. Simulations of bioelectrical activity were conducted using openCARP. In the absence of sinus pacing, simulations were run for 1000 ms to assess the likelihood of ectopic beats being about to outcompete sinus pacing. Point endocardial stimuli (3×) were delivered every 1000 ms to mimic 60 bpm sinus rhythm in simulations that were run for 2500 ms.

Results, Conclusions, and Discussions: In the absence of sinus rhythm, all spontaneous beating rates were associated with at least one propagating ectopic beat for graft patterns 1 and 2. Graft patterns 3-6 had a lower overall occurrence of ectopic beats regardless of the intrinsic beating rate of the hPSC-CM model used. Overall, faster hPSC-CM spontaneous beating rate leads to an increased burden of ectopic excitation (Fig C). With simulated sinus activity applied (i.e., non-hPSC-CM excitations originating near the endocardial surface, representing activity from the Purkinje system), the number of observed ectopic beats decreased but at least one example of focal arrhythmia was observed for all but the slowest-beating hPSC-CM cell scale model (1.1 Hz; Fig D). Our results indicate that to outcompete sinus rhythm and sustain robust, high rate focal VT, the spontaneous beating rate of engrafted hPSC-CM must be faster than the observed cell excitation frequencies in vitro (i.e., data used to calibrate many published cell-scale models). Even when sinus pacing is added, faster-beating hPSC-CMs create an electrical milieu conducive to focal VT. This is the first in silico work to successfully recapitulate the focal VT phenomenon in a biophysically detailed and geometrically realistic hPSC-CM injection model. More work is needed to derive the electrocardiographic signatures of these focal VTs and to expand this framework to a larger number of human geometries.

Acknowledgements (Optional):