Virtual Heart Models Reveal Why Some Patients Develop Heart Block After TAVR

Transcatheter aortic valve replacement, or TAVR, has transformed the treatment of aortic stenosis, offering patients a minimally invasive alternative to open-heart surgery. Yet the procedure carries a stubborn complication that has puzzled cardiologists for years: damage to the heart’s electrical wiring system. The most common of these injuries is left bundle branch block, a disturbance in the conduction pathway that coordinates contraction of the heart’s main pumping chamber. Now, a team of biomedical engineers and cardiologists has built patient-specific computer simulations that quantify, for the first time in bicuspid aortic valve patients, exactly how much mechanical stretch the conduction tissue endures during valve implantation—and the results point to a strain threshold that could one day help physicians predict who is at risk before the catheter ever enters the body.

The study, published in the Annals of Biomedical Engineering by Salwa Anam, Brandon Kovarovic, and colleagues at Stony Brook University working with collaborators at Rabin Medical Center and Tel Aviv University, focused on a particularly challenging patient population. People with bicuspid aortic valve, a congenital condition in which the valve has only two leaflets instead of three, are known to face elevated rates of conduction abnormalities after TAVR. Their unusual valve anatomy and calcification patterns make device deployment less predictable, and prior computational studies had evaluated generalized conduction risk without isolating the specific mechanics of left bundle branch block or comparing different valve designs in this group.

To address that gap, the researchers analyzed cardiac CT scans from six patients with type I bicuspid aortic valve who had undergone TAVR. Two of the patients developed left bundle branch block after the procedure, while four experienced no conduction abnormalities at all. From those scans, the team constructed detailed finite element models that captured each patient’s individual anatomy, including the distribution and geometry of calcium deposits on the valve leaflets—features that dramatically alter how a transcatheter valve expands and presses against surrounding tissue.

A central challenge was defining where the vulnerable conduction tissue actually sits. The left bundle branch runs along the left side of the interventricular septum, close to the membranous septum, a thin fibrous region of the heart’s wall. The team used the membranous septum length and established anatomical landmarks to delineate a left bundle branch-adjacent region in each model, allowing them to compute mechanical metrics specifically in the zone where implantation trauma is most likely to occur. This anatomically grounded approach distinguishes the work from earlier simulations that treated the risk region more coarsely.

The simulations reproduced the implantation of deployed self-expanding devices in each patient and also placed virtual balloon-expandable devices at a standardized implantation depth. From these virtual procedures, the researchers calculated area-weighted logarithmic strain, contact pressure, oversizing index, eccentricity index, and expansion metrics throughout the conduction-adjacent tissue. Logarithmic strain, a measure that captures large deformations accurately, served as the primary indicator of how severely the tissue surrounding the conduction pathway was stretched by the expanding valve stent.

The separation between the two patient groups was striking. In the patients who developed left bundle branch block, strain values in the left bundle branch-adjacent region ranged from 0.15 to 0.24. In the patients without any conduction abnormality, the corresponding values ranged from just 0.03 to 0.11. The two distributions did not overlap, and the team identified 0.11 as a preliminary cohort-specific separation value—any patient whose simulated strain exceeded that level fell into the injured group. While the finding comes from only six patients and must be validated in larger cohorts, the clean separation suggests that tissue stretch, rather than some other mechanical factor, may be the dominant driver of this specific injury.

Equally revealing was what did not predict the outcome. Contact pressure between the valve stent and the surrounding tissue showed no consistent association with whether a patient developed left bundle branch block. This matters because pressure has been a focus of earlier computational work on conduction abnormalities, including patient-specific studies of self-expanding valves that examined the role of contact pressure in new conduction disturbances. The new results suggest that strain differences reflect multiple contributing factors—anatomy, calcification, device geometry, and deployment depth acting together—rather than a single compressive force, and that pressure alone is an incomplete proxy for injury risk.

The device comparison added another layer of practical insight. Across the cohort, mean strain in the left bundle branch-adjacent region was higher with self-expanding devices than with balloon-expandable ones, at 0.11 versus 0.08. Self-expanding valves exert continuous radial force and can conform asymmetrically to eccentric, calcified anatomy, while balloon-expandable valves expand more uniformly at a fixed diameter. The trend hints that device selection could meaningfully modulate conduction risk in bicuspid patients, though the authors are careful to note that these are risk trends observed in a small proof-of-concept cohort, not definitive guidance for clinical decision-making.

The methodological rigor behind the models reflects years of prior validation work by the same group, which has previously validated patient-specific structural and flow simulations against actual transcatheter bicuspid valve replacement procedures, and has characterized the mechanical properties of excised calcified aortic leaflet tissue to improve the fidelity of in silico modeling. Incorporating realistic leaflet calcification is not a cosmetic detail: calcified nodules stiffen the landing zone, redirect stent expansion, and can concentrate deformation in precisely the regions where the conduction system travels. By embedding patient-specific calcification into the finite element framework, the simulations capture mechanical interactions that idealized models miss entirely.

The clinical stakes are considerable. New left bundle branch block after TAVR has been linked in registry and cohort studies to adverse outcomes, including progression to cardiomyopathy, pacemaker dependence, and changes in cardiac remodeling, and conduction disturbances remain among the most frequent complications of the procedure overall. A validated computational tool that estimates strain-based risk before implantation could allow heart teams to compare devices virtually, adjust implantation depth, or flag patients who warrant closer post-procedural monitoring. The researchers themselves emphasize the caveats: larger cohorts and sensitivity analyses are needed before the 0.11 strain threshold can be used for risk stratification or device selection in bicuspid aortic valve TAVR. But as proof of concept, the study demonstrates that the answer to a common clinical mystery may lie in a number that a computer can compute before the procedure begins—turning the heart’s electrical vulnerability into a measurable, and potentially avoidable, mechanical quantity.

Subject of Research: Computational modeling of mechanical strain-induced left bundle branch block after transcatheter aortic valve replacement in bicuspid aortic valve patients

Article Title: Strain-Based Assessment of Post-TAVR Left Bundle Branch Block in Bicuspid Aortic Valve Using Patient-Specific Computational Modeling

Article References: Strain-Based Assessment of Post-TAVR Left Bundle Branch Block in Bicuspid Aortic Valve Using Patient-Specific Computational Modeling. (n.d.). https://doi.org/10.1007/s10439-026-04399-w

Image Credits: AI Generated

DOI: 10.1007/s10439-026-04399-w

Keywords: TAVR, bicuspid aortic valve, left bundle branch block, cardiac conduction abnormality, finite element modeling, patient-specific simulation, logarithmic strain, contact pressure, self-expanding valve, balloon-expandable valve, leaflet calcification, biomechanics