Non-invasive spinal electro-magnetic stimulation (SEMS): a tool for evaluation and modulation of lower limb spinal-muscular transmission in healthy adults

Objective Our earlier electrophysiological recordings using animal models revealed diminished transmission through spared fibers to motoneurons and leg muscles after incomplete spinal cord injury (SCI). Administration of spinal electro-magnetic stimulation (SEMS) at specific parameters induced transient improvement of transmission at neuro-muscular circuitry in SCI animals. In the current human study, we sought translate this knowledge to establish optimal parameters of SEMS for (i) neurophysiological evaluation via Compound Motor Action Potential (CMAP); and (ii) modulation at neuro-muscular circuitry via H-reflex and M-wave response in 12 healthy adults. Methods SEMS application was with a coil positioned over T12-S1 spinal levels. SEMS-evoked CMAP-responses were wirelessly measured simultaneously from biceps femoris (BF), semitendinosus (ST), vastus lateralis (VL), soleus (SOL), medial gastrocnemius (MG) and lateral gastrocnemius (LG) muscles. We also examined effects of SEMS trains on H-reflex and M-wave responses. H-reflexes and M-waves were measured simultaneously from SOL, MG and LG muscles and evoked by peripheral electrical stimulation of tibial nerves before and after each SEMS session. Results Spinal levels for SEMS application to evoke CMAP-responses in corresponding muscles and amplitude/latency of these responses have been established. SEMS applied over L4-S1 spinal levels at 0.2 Hz rate for 30 min induced facilitation of H-reflexes and M-responses. Facilitation lasted for at least 1 hour after stopping SEMS and was associated with a decrease in threshold intensity and leftward shift of recruitment curve for H-reflex and M-wave. SEMS did not alter TMS-evoked responses in hand muscles. Conclusion SEMS is a novel, non-invasive approach for sustained neuromodulation of H-reflex and M-wave responses in triceps surae muscle group. The parameters of SEMS application established in this study for evaluation and neuromodulation of neural pathways innervating leg muscles in healthy individuals may be used as a reference for neurophysiological evaluation and long-lasting plasticity of the lower limb spino-neuromuscular circuitry in individuals with SCI.


Introduction
Electro-magnetic stimulation is based on the principle of electromagnetic induction of an 70 electric field through intact tissue to underlying structures systems [16][17]. Transcranial electro-71 magnetic stimulation (TMS), since its initial introduction about three decades ago [18], has been 72 widely used for diagnostic applications and repetitive TMS was found to alter excitability at 73 cortico-motor circuitry [17,[19][20]. Responses evoked by TMS, however, are reliably recorded 74 from arm but not leg muscles, particularly in human participants with neurological impairments, 75 including spinal cord injury [21]. Consistent with these reported results of human studies, our 76 recent experiments using animal models revealed that TMS-evoked responses could be reliably 77 recorded from the hind limb muscles only in the naïve animals, but not in adult rats with SCI [22]. 78 However, responses evoked by electro-magnetic stimulation at spinal thoracic and lumbar levels 79 could be reliably recorded from hind limb muscles in rats with chronic SCI [1]. In fact, EMG    intensity were determined in their first session. To accomplish this, a total of 6-10 single pulses 205 were delivered to each spinal level from T12 to S1 and the stimulus intensity with 10% increments 206 ranged from 40% to 80% was applied until EMG recordings from the greatest number of muscles 207 were visible. After the optimal spinal level and SEMS intensity were determined, they were used 208 to deliver a 30 minutes of single pulse SEMS that was applied repetitively with 0.  Where S is the stimulus intensity, Hmax is maximum amplitude of H-reflex. S50 is stimulus 231 intensity needed to obtain 50% of Hmax, and b is the steepness of the curve. In addition, we 232 calculated the slope of the curve at S50 using following equation

278
With the same stimulation intensity, the optimal spinal level to evoke CMAP responses in 279 the triceps surae muscle group (MG, LG, SOL) was the L5-S1 spinal levels. The mean amplitude 280 for the SOL muscle at 70% of the maximum coil intensity was 0.15 ± 0.02 mV whereas the MG 281 and LG muscles showed 0.14 ± 0.3 mV and 0.10 ± 0.02 mV amplitude respectively (n=7). For 282 the VL muscle, the optimal spinal level was T12-L1. The mean amplitude for the VL muscle at 283 70% of maximum coil intensity was 0.14 ± 0.07 mV. For the BF and ST muscles, the optimal 284 spinal stimulation level was L1-L3. The mean amplitude for BF muscle at 70% of maximum coil 285 intensity was 0.03 ± 0.01 mV and ST was 0.05 ± 0.01 mV respectively.
Latencies of CMAP responses were also different depending on the muscle. Latency for 287 the VL muscle responses evoked from L1 SEMS was 11.8 ± 1.1 ms, 13.5 ± 1.3 ms for BF muscle    curve for both H-and M-responses; however, this shift was different among participants (Fig 4).       In order to understand whether SEMS induced changes in H-reflex and in M-wave are comparable, 405 we have also analyzed correlation between SEMS-induced changes in the parameters of H-reflex 406 and M-wave recruitment curves. As shown in Table 3,

Modulation of H-reflex by SEMS train is long-lasting and sustained after stop
413 of SEMS 414 We have also examined whether the effect of SEMS are long-lasting, i.e. how long the 415 observed threshold changes were sustained by measuring H-reflex before, immediately after 416 SEMS and after 1 hour post stopping of SEMS (Fig 6). The threshold intensity required for H 417 reflex was still significantly lower after 1 hour post SEMS administration;79.8 ± 3.3 % post-SEMS 418 compared to 100% pre-SEMS (n=4; p<0.05). Importantly, the decrease in the MT intensity 419 following SEMS application, however, was not associated with the changes in Hmax. Hmax   429 We have also examined whether SEMS administration may affect TMS induced responses. 430 We recorded TMS evoked responses from hand FDI muscle before and after SEMS administration.

431
Representative traces recorded from FDI muscle before and after SEMS in the same participant 432 presented in Fig 7A. We did not observe any changes in TMS evoked responses after SEMS.