Modeling Cardiac Effects of Vagus Nerve Stimulation in Heart Transplant Recipient

Heart transplantation (HTx) is the last resort for patients with severe heart failure, but an inevitable consequence of the surgery is cardiac denervation. This leads to a markedly increased resting heart rate and the inability of the heart to adapt to physical activity, thus reducing patient quality of life. Today, neuromodulation techniques such as vagus nerve stimulation (VNS) are successfully applied to treat various conditions including Parkinson disease, treatment-resistant depression, or epilepsy. As the vagus nerve (VN) has several projections to the heart, it plays a crucial role in chronotropic-, dromotropic- and inotropic control. Therefore, VNS may be a promising technique to artificially restore vagal cardiac control in those suffering from cardiac denervation. Although VNS is typically applied in an asynchronous fashion (A-VNS), for cardiac control it can be synchronized with respect to the cardiac cycle (S-VNS). This is often done by triggering a burst of impulses with the R-peak of the ECG. However, to this date, the cardiac effects of VNS are not fully understood, thus complicating the development of effective stimulation paradigms to restore cardiac control. To further investigate the effects of VNS, in this thesis, a numerical model of the human cardiovascular system was extended to integrate a lumped parameter model of the VN that is target to extracellular stimulation. The model consists of (1) the hemodynamic system, comprising the heart and vessels; (2) the autonomic sympathetic cardiac control; (3) the sinoatrial node; (4) the VN realized as Hodgkin-Huxley-type multi-axon model; and (5) the VN terminals represented by a three-compartment acetylcholine release model. Moreover, a stimulator was integrated into the model as a pulse generator that allows eliciting action potentials within the VN axons by extracellular stimulation through a monopolar point electrode. The pulse generator is configurable allowing the manipulation of stimulation parameters including intensity (I), pulse width (PW), frequency (F), number of pulses (NP), and trigger delay (D). The model was then tuned based on published data on the cardiac effects of VNS obtained from in-vivo experiments in sheep. To investigate the influence of the stimulation parameters on cardiac effects they underwent a global sensitivity analysis (GSA). The GSA was performed for A- and S-VNS, in both, the model of healthy individuals and HTx recipients. The results show that for S-VNS I has the greatest influence, followed by PW and NP, while F and D play a minor role. The results for A-VNS are similar as in S-VNS, however, F has a markedly increased influence on the cardiac effect. For both A-VNS and S-VNS, cardiac denervation following HTx shows no significant differences for chronotropism and dromotropism while an overall reduction for inotropism was found. Finally, the model was validated, comparing the results to literature data, different from that used for tuning. The model was found capable to reproduce the chronotropic and dromotropic effects of VNS with good accuracy. However, for the inotropic response, the model predictions are in conflict with the findings in literature. Overall, the study could help to identify potential targets stimulation parameters for the design of a closed-loop control strategy. Moreover, the model forms a viable foundation to further investigate the cardiac effects of VNS on the denervated heart.