Abstract
A goal in modern pacemaker technology is to adapt the timing of stimulation to the metabolic demand and to optimize the hemodynamics of the diseased heart. Because the plasma concentration of mixed venous oxygen is a good indicator of workload and circulation, we developed an implantable coulometric blood oxygen sensor to improve the rate adaptation. Two different working electrode (WE) materials in direct contact with the blood were tested, smooth glassy carbon and gold. Two reference electrode (RE) concepts using Ag/AgCl and one made of porous pyrolytic carbon were evaluated. The used counter electrode (CE) consisted of the titanium housing of the pulse generator partly coated with carbon.
To perform chronic in vivo studies an implantable pacemaker system with electronics for chronocoulometric oxygen detection was developed. Both single and double potential step techniques were evaluated. The hemispherical or cylindrical working electrode was placed on the stimulation lead inside the heart. Potential steps were periodically synchronized to end-diastole, when the heart is filled with blood. The reduction potential was 1V for carbon and 0.8V for gold and the duration of the imposed overpotential 15ms, while the electrode potential was floating in-between. The electrical current that is limited by oxygen diffusion and reaction rate is a function of the blood oxygen concentration and used for calculation and regulation of the stimulation rate.
Ellipsometric in vitro studies on human blood protein adsorption to the two materials, pyrolytic carbon and gold were performed to both qualitatively and quantitatively identify the initial interactions with the sensor materials at electrochemical overpotentials.
Studies on 31 animals were performed to refine and evaluate long-term in vivo stability and biocompatibility. Standard transvenous lead implantation technique was used. To create a realistic animal model of a pacemaker patient, the AV node of 5 dogs was destroyed by RF-ablation. The sensor stability and response to exercise was followed monthly. Post-mortem examinations of the electrode surfaces and tissue response were performed. The maximum implantation time was 4.5 years.
The results show that the sensor stability and response to workload was excellent during the study time. Adsorbed plasma proteins on the implanted gold surface did not decrease oxygen transport or reaction efficacy. Due to dislocation and mechanical irritation 3 electrodes grew into the endocardium and became insensitive. This complication can probably be avoided by a slight redesign of the sensor lead. No adverse tissue reactions have been observed at the investigated working electrode materials.
