The refractory period of fast conducting corticospinal tract axons in man and its implications for intraoperative monitoring of motor evoked potentials

Abstract

Objective: To determine the absolute and relative refractory period (RRP) of fast conducting axons of the corticospinal tract in response to paired high intensity (HI or supramaximal) and moderate intensity (MI or submaximal) electrical stimuli. The importance of the refractory period of fast conducting corticospinal tract axons has to be considered if repetitive transcranial electrical stimulation (TES) is to be effective for eliciting motor evoked potentials (MEPs) intraoperatively.

Methods: Direct (D) waves were recorded from the epidural space of the spinal cord in 14 patients, undergoing surgical correction of spinal deformities. To assess the absolute and RRPs of the corticospinal tract, paired transcranial electrical stimuli at interstimulus intervals (ISI) from 0.7 to 4.1 ms were applied. Recovery of conditioned D wave at short (2 ms) and long (4 ms) ISI was correlated with muscle MEP threshold. The refractory period for peripheral nerve was tested in comparison to that for the corticospinal tract. In four healthy subjects sensory nerve action potentials of the median nerve were studied after stimulation with paired stimuli.

Results: HI TES revealed a mean duration of 0.82 ms for the absolute refractory period of the corticospinal tract, while MI stimulation resulted in a mean refractory period duration of 1.47 ms. Stimuli of HI produced faster recovery of D wave amplitude during the RRP. Furthermore, short trains of transcranial electrical stimuli did not elicit MEPs when D wave showed incomplete recovery. A similar influence of stimulus intensity on recovery time was found for the refractory period of peripheral nerve.

Conclusions: The recovery of D wave amplitude is dependent upon stimulus intensity. High intensity produces fast recovery. This is an important factor for the generation of MEPs. When HI TES is used to elicit MEPs, short and long ISIs are equally effective. When MI TES is used to elicit MEPs, only a long ISI of 4 ms is effective.

Introduction

Monitoring of motor evoked potentials (MEPs) during spine and spinal cord surgery has become a valuable method for providing continuous information about the functional integrity of a patient's motor system. Transcranial electrical stimulation (TES) has been proven feasible for intraoperative use in the monitoring of MEPs from limb muscles (mMEPs). Short trains of transcranial stimuli, used in patients under general anesthesia, can reliably elicit mMEPs. (Pechstein et al., 1994, Deletis et al., 1995, Pechstein et al., 1996, Jones et al., 1996). However, factors such as preexisting neurological deficits, previous radiation therapy, and previous surgery are unfavorable when attempting to elicit MEPs, and may prevent continuous MEP monitoring (Morota et al., 1997, Kothbauer et al., 1998). The anatomical location and extent of pathology might also influence the efficacy of this stimulation technique. The optimal parameters for TES of motor pathways are still to be determined.

Short trains of TES expose the corticospinal tract to repetitive stimuli with an interstimulus interval (ISI) of 2–5 ms. Repetitive activation of corticospinal tract fibers is dependent on their refractory period. The travelling waves in the corticospinal tract induced by TES can be recorded via epidural electrodes. Responses consisting of a complex of direct and indirect waves (D and I waves) are considered to be orthodromic electrical activity recorded from fast conducting corticospinal tract neurons (Patton and Amassian, 1954, Boyd et al., 1986, Rothwell et al., 1994). In primates, I waves reflect transsynaptically activated semisynchronous activity of the corticospinal tract in response to electrical (Patton and Amassian, 1954) and particularly oriented magnetic (Amassian et al., 1989) stimulation. I waves are significantly suppressed by anesthetic agents (Hicks et al., 1992).

Deletis et al. reported on recovery of D wave under the condition of general anesthesia. Using a paired stimulus technique (delivering a conditioning and test stimulus), the amplitude and latency of the conditioned D wave were analyzed to determine the electrophysiological properties of the corticospinal tract during the relative refractory period (RRP). The authors have reported that complete recovery of human corticospinal tract from the RRP occurred approximately 4 ms after the conditioning stimulus, indicating that 4 ms is an optimal ISI in transcranial train stimulation for generation of mMEPs (Deletis et al., 2001a). However, the refractory period was only assessed by stimuli of submaximal intensity in this study. The current study was intended to complement previously reported data on corticospinal tract refractory period by testing the influence of stimulus intensity on recovery of D wave amplitude and latency at various ISI.

Corticospinal tract and peripheral nerve axons share properties of recovery from excitation after a conditioning discharge (Kiernan et al., 1996, Burke et al., 2000, Chan et al., 2002). Thus the median nerve can serve as a feasible peripheral model for observations on corticospinal tract refractoriness. To correlate data obtained from fast conducting axons of the corticospinal tract with peripheral nerve axons, a protocol similar to that used to assess the corticospinal tract refractory period was applied to sensory fibers of the median nerve.

Critical data derived from studies of the refractory period of the corticospinal tract might provide valuable information for optimizing the parameters used to intraoperatively elicit mMEPs with trains of TES. However, translation of TES mediated corticospinal tract activity into generation of mMEPs is not a linear process. Generation of mMEPs is not solely dependent on generation of D waves by TES. Additional interactions at the level of the cortical motor neuron (e.g. facilitation of I waves), as well as at the level of the alpha motor neuron (e.g. temporal summation of corticospinal tract volleys) might influence the generation of mMEPs (Patton and Amassian, 1954, Amassian et al., 1987, Sloan, 2002, Taylor et al., 1993, Deletis and Kothbauer, 1998, Deletis et al., 2001b). If stimulus intensity determined D wave recovery, then generation of mMEPs by trains of TES would be influenced accordingly. Reported data on the optimal ISI for repetitive TES vary (Kalkman et al., 1995, Pechstein et al., 1996, de Haan et al., 1998, Deletis et al., 2001a, MacDonald, 2003). One reason for discrepancies may be that a correlation between ‘optimal ISI’ and stimulus intensity has not yet been established. In order to test the hypothesis that D wave recovery directly influences the ability to elicit mMEPs by trains of stimuli we have recorded D wave and mMEPs simultaneously after applying HI (high intensity or supramaximal) and MI (moderate intensity or submaximal) TES.