Effects of expiratory muscle activation via high-frequency spinal cord stimulation

Abstract

In persons with spinal cord injury, lower thoracic low-frequency spinal cord stimulation (LF-SCS; 50 Hz, 15 mA) is a useful method to restore an effective cough. Unfortunately, the high-stimulus-amplitude requirements and potential activation of pain fibers significantly limit this application in persons with intact sensation. In this study, the mechanism of the expiratory muscle activation, via high-frequency SCS (HF-SCS; 500 Hz, 1 mA) was evaluated in dogs. In group 1, the effects of electrode placement on airway pressure generation (P) was evaluated. Maximal P occurred at the T9-T10 level with progressive decrements in P at more rostral and caudal levels for both LF-SCS and HF-SCS. In group 2, electromyographic (EMG) latencies of internal intercostal muscle (II) activation were evaluated before and after spinal root section and during direct motor root stimulation. Onset time of II EMG activity during HF-SCS was significantly longer (3.84 ± 1.16 ms) than obtained during direct motor root activation (1.61 ± 0.10 ms). In group 3, P and external oblique (EO) EMG activity, before and after sequential spinal section at the T11-T12 level, were evaluated. Bilateral dorsal column section significantly reduced EO EMG activity below the section and resulted in a substantial fall in P. Subsequent lateral funiculi section completely abolished those activities and resulted in further reductions in P. We conclude that 1) activation of the expiratory muscles via HF-SCS is dependent entirely on synaptic spinal cord pathways, and 2) HF-SCS at the T9 level produces a comparable level of muscle activation with that achieved with LF-SCS but with much lower stimulus amplitudes. NEW & NOTEWORTHY The findings in the present study suggest that lower thoracic high-frequency spinal cord stimulation with low stimulus currents results in sufficient activation of the expiratory muscles via spinal circuitry to produce large positive airway pressures sufficient to generate an effective cough mechanism. This method, therefore, may be applied in patient populations with intact sensation such as stroke and amyotrophic lateral sclerosis to restore an effective cough.