Effects of subcutaneous nerve stimulation with blindly inserted electrodes on ventricular rate control in a canine model of persistent atrial fibrillation

Abstract

Subcutaneous nerve stimulation (ScNS) delivered directly to large subcutaneous nerves can be either antiarrhythmic or proarrhythmic, depending on the stimulus output. The purpose of this study was to perform a prospective randomized study in a canine model of persistent AF to test the hypothesis that high-output ScNS using blindly inserted subcutaneous electrodes can reduce ventricular rate (VR) during persistent atrial fibrillation (AF) whereas low-output ScNS would have opposite effects. We prospectively randomized 16 male and 15 female dogs with sustained AF (>48 hours) induced by rapid atrial pacing into 3 groups (sham, 0.25 mA, 3.5 mA) for 4 weeks of ScNS (10 Hz, alternating 20-seconds ON and 60-seconds OFF). ScNS at 3.5 mA, but not 0.25 mA or sham, significantly reduced VR and stellate ganglion nerve activity (SGNA), leading to improvement of left ventricular ejection fraction (LVEF). No differences were found between the 0.25-mA and sham groups. Histologic studies showed a significant reduction of bilateral atrial fibrosis in the 3.5-mA group compared with sham controls. Only 3.5-mA ScNS had significant fibrosis in bilateral stellate ganglions. The growth-associated protein 43 (GAP43) staining of stellate ganglions indicated the suppression of GAP43 protein expression in the 3.5-mA group. There were no significant differences of nerve sprouting among all groups. There was no interaction between sex and ScNS effects on reduction of VR and SGNA, LVEF improvement, or results of histologic studies. We conclude that 3.5-mA ScNS with blindly inserted electrodes can improve VR control, reduce atrial fibrosis, and partially improve LVEF in a canine model of persistent AF.