Abbreviations

ВSCI — spinal cord injury

SC — spinal cord

SS — spinal shock

SSEP — somatosensory evoked potentials

ESG — electrospinogram

CPC — constant potential capacity

TMS — transcranial magnetic stimulation

TES — transcranial electrostimulation

IONM — intraoperative neurophysiological monitoring

ENMG — electroneuromyography

NMG — neuromyography

The incidence of spinal cord injury worldwide varies considerably and ranges from 8 to 58 cases per 1 million [1]. Spinal cord injury accounts 3—5% of blunt injuries and 5.5 — 17.8% of all musculoskeletal injuries [2]. Reversibility of neurological deficit associated with SCI depends on damage to neural structures and severity of complications (hemorrhage and secondary spinal cord ischemia). Clinical severity of SCI is assessed in accordance with the classification proposed by the American Spinal Injury Association (ASIA) (1992). The last one is based on modified scale of H.L. Frankel et al. (1969) [3, 4]. According to this scale, victims are divided into 5 groups (A, B, C, D, E) depending on severity of motor and sensitive disorders arising after SCI. ASIA scoring system requires adequate speech contact with a victim. Therefore, this approach is not always advisable in patients with concomitant traumatic brain injury (TBI), speech disorders or drug-induced depression of consciousness. Another factor complicating prediction of SCI outcome may be spinal shock. Indeed, SS creates a picture of complete SCI in acute post-traumatic period, although full functional recovery may occur later [5, 6]. Differential diagnosis of functional and anatomical SCI is very important for the choice of treatment strategy especially for choosing the type and timing of surgery [7].

Instrumental diagnosis of SCI is based on neuroimaging (multispiral computed tomography and magnetic resonance imaging in various modifications). However, unambiguous prediction of the outcome of trauma based on MR- and CT-images is possible only in patients with complete interruption of spinal cord. In some cases, symptoms do not correspond to neuroimaging data or their evaluation is impossible due to above-mentioned reasons. Functional assessment of conductors and gray matter is necessary in these patients [2]. For this purpose, various methods based on electrophysiological evaluation of spontaneous and evoked bioelectrical potentials of neurons and muscles are used. These methods are united under the common name "neurophysiological".

Objective. To determine the capabilities of neurophysiological survey for prediction of the outcome of spinal cord injury.

Functional assessment of the spinal cord

Neurophysiological methods for diagnosis of SCI and functional analysis of SC may be divided into two main groups: registration of spontaneous bioelectrical spinal activity and assessment of action potential of muscle or nerve tissue in response to central or peripheral neurostimulation.

Methods for recording the spontaneous bioelectrical spinal activity. Historically, the first method for functional survey of SC was registration of spontaneous bioelectrical activity (electrospinogram) and estimation of constant potential capacity. The possibility of ESG registration in human was first reported by J.Pool in 1946. Initially, this approach implied epidural insertion of a special electrode using lumbar puncture [8]. Technical possibility of ESG registration by skin electrodes attached within spinal cord projection is reported in various publications [9, 10]. The technique consists in evaluation of bioelectrical activity of spinal cord with signal amplification and filtering parameters similar to those in electroencephalography (EEG). Analysis of slow cyclic changes of constant spinal potential simultaneously with EEG is possible if registration of ultra-low frequency oscillations is technically possible (<0.01 Hz). ESG and CPC of spinal cord reflect spontaneous bioelectrical activity generated by axon membranes and probably motor neurons of anterior horns [11]. These methods are only able to confirm intact structure and possibly metabolic activity of spinal cord. However, conductive function cannot be analyzed.

Registration of evoked spinal activity. Functional analysis of the axon involved into both motor and sensory activity requires registration of action potential arising in its distal segment in response to proximal stimulation. These methods of neurophysiological diagnosis are commonly determined as evoked potentials (EP). They are divided into two main groups — motor and sensory EP.

Analysis of motor EPs (MEPs) implies stimulation of pre-central gyrus and parameters of muscle contractility 15—50 ms later. Transcranial magnetic stimulation, transcranial electric stimulation and direct cortical electrical stimulation are used. However, the last two approaches are only possible under anesthesia due to significant painfulness. Moreover, direct cortical stimulation also requires craniotomy with dura mater dissection in projection of precentral gyrus. Another type of motor EP is D-wave technique [12]. In this case, TES is followed by registration of response by a special invasive electrode. This electrode is inserted into epidural or subdural space over dorsal surface of spinal cord after laminectomy. Motor EPs in response to TMS are usually studied in patients with SCI while TES and D-wave are only performed during intraoperative neurophysiological monitoring (IONM) [13, 14].

Analysis of sensory EPs implies stimulation of peripheral segments of analyzers, sensitive or mixed nerves [15]. Somatosensory EPs in response to stimulation of tibial or median nerves are used in victims with SCI. Recording electrode is placed on the scalp in projection of functional area of the arm in case of median nerve stimulation and on the vertex in case of tibial nerve stimulation. Additional electrodes are placed within cervical and lumbar thickening, as well as over large plexuses of peripheral nerves. Analysis of time difference of EP from sequentially located neural structures and their amplitude is valuable to localize pathological process and assume its mechanism (demyelination or axonal damage) [16].

Motor and somatosensory EPs correlate with functional state of different conductors (lateral and posterior funiculi, respectively). Therefore, analysis of both modalities of EPs is required for complete assessment of spinal cord function.

Neurophysiological survey of the peripheral nervous system

Myography is another method of neurophysiological diagnosis of SCI. Myography data may be valuable for indirect assessment of the course of spinal shock and rehabilitation potential of SCI although this approach is used for functional survey of only peripheral nervous system. Myography includes a significant number of different techniques. The main ones are electroneuromyography (ENMG, Nerve Conduction Study) and neuromyography performed by bipolar concentric needle electrode (NMG, myography) [17].

ENMG consists in electrical stimulation of the nerve and subsequent evaluation of the total action potential of the muscle (M-response) or nerve (S-response). This method allows estimating severity of axonal damage of the second motor neuron or preganglionary segment of sensory nerve. One of the types of ENMG is F-wave analysis. F-wave is a low amplitude polymorphic muscle oscillation occurring in a short interval after M-response. F-wave genesis is well studied. It is the result of excitation of motor neurons of anterior spinal horns by electric impulse spreading in proximal direction [18].

Needle NMG data reflect the processes of muscle fiber rearrangement associated with impaired innervation regardless the cause of impairment. Acuteness, stage and severity of this process may be estimated using NMG. However, it is an invasive technique requiring stable contact with a patient for arbitrary contraction of muscles with various intensity [19].

Informative value of neurophysiological survey in prediction of SCI

Various retrospective trials and systematic reviews are devoted to analysis of informative value of ESG, CPC, SSEP, TMS and ENMG.

ESG and CPC reflecting spontaneous axonal activity were the first instrumental methods for assessment of SC damage and prediction of the outcomes of SCI. In 1946, J.L. Pool described the technique of ESG recording in human and registered acute potential outbreaks in a patient with paraplegia. The author related their occurrence with spontaneous attacks of spastic leg muscle tension and suggested that such spinal bioelectrical activity should be called "spinal epilepsy" [8]. The majority of researches devoted to ESG and CPC registration are experimental [11, 20]. These methods have shown high informative value for assessment of spinal cord damage caused by trauma or ischemia [11]. However, these methods have not received wide practical application because they are not informative for prognosis of SCI. Indeed, SC damage may be easily established by clinical examination and neuroimaging while intact bioelectrical spinal activity below the area of spinal interruption does not allow making conclusions with high degree of reliability.

K. Shimoji and et al. [21] first described technical possibility of registration of ESG changes in response to electrical stimulation of peripheral nerves in humans in 1971. Subsequently, the technique was supplemented by digital processing of signal for accumulation and averaging and called ESG [15]. In patients with SCI, the method of SSEPs allows stimulation of sensory segment of peripheral nerve innervating certain dermatome below SCI while the response is recorded from upper spinal or cerebral regions. Registration of evoked potentials from the leads in projection of functional cortex of limbs and above spinal cord thickening indicates intact conduction in posterior funiculi and their functional integrity.

Informative value of SSEPs for predicting the outcomes of SCI was studied in several studies and summarized in a systematic review of P.K. Bedi in 2015 [22]. For example, there was similar prognosis of walking restoration after SCI in case of clinical assessment with ASIA impairment scale and instrumental assessment with SSEPs registration [23]. M. Spiess et al. (2015) studied SSEPs during tibial nerve stimulation in 297 patients with SCI within the first year after injury. The authors showed that P38 peak amplitude changes over this time in 10% of patients. Importantly, initial absence of cortical SSEPs was not an absolute predictor of unfavorable prognosis of walking recovery. These data may be due to the fact that SSEPs reflect exclusively functional state of posterior spinal funiculi. These structures are composed of conductors of deep sensitivity and do not allow estimating the efferent conduction from neurons of pre-central cerebral gyrus to the limb muscles.

Clinical introduction of TMS for registration of motor evoked potentials ensured survey of functional integrity of the corticospinal tract in patients with SCI without sedation. It is a method of painless and safe non-invasive highly intensive transcranial stimulation of the primary motor cortex [24, 25]. A. Curt et al. reported similar informative value of TMS and ASIA scale for predicting upper limb function recovery after SCI [27]. Moreover, analysis of the amplitude of evoked motor responses and their latency may be valuable in patients with spinal cord contusion to estimate severity of nerve tissue damage although these data are not informative for predicting recovery [28].

Diagnostic TMS implies recording of evoked motor potentials from the limb muscles. Therefore, function of peripheral nerves and motor neurons of anterior spinal horns should be preserved for correct interpretation of data. In patients with SCI, injury of peripheral motor neurons may be caused by direct traumatic effect and secondary spinal cord ischemia [28]. These processes result wallerian degeneration of axons and their gradual destruction followed by muscle atrophy. As a result, TMS becomes impossible for diagnosis of severity of spinal cord injury.

Myography is used for functional diagnosis of peripheral nervous system. Reduced amplitude of M-responses (ENMG) and spontaneous activity in needle NMG indicate axonal lesion. However, these methods are not informative for differential diagnosis of radicular and spinal cord lesion in case of complicated SCI [29].

SS is another cause reducing informative value of TMS-MEP and SSEPs in acute period of SCI. This is a reversible functional blockade of spinal cord structures below the damage area. Clinical severity of SS depends on spinal cord damage dynamics [7]. Diagnosis of spinal shock is associated with considerable difficulties due to similar clinical symptoms of anatomical and functional spinal cord interruption especially in case of ambiguous neuroimaging data [30]. Pathophysiological mechanism of SS development is determined by rapid extracellular release of potassium ions from intracellular space. This process is a result of direct spinal cord trauma and vascular injury followed by hemorrhage and ischemia [7]. Biochemical cascade leads to downward retardation that is clinically manifested by transient disappearance of tendon and vegetative reflexes and lost sensitivity [31]. C.D. Barnes et al. (1962) experimentally found that electrographic biomarker of these processes is hyperpolarization of spinal motor neurons from 2 to 8 mV [32]. Analysis of F-wave is one of the stages of ENMG. This method allows calculation of nerve impulse propagation velocity along the proximal segment of peripheral nerve and functional estimation of the motor neuron [19]. D.S. Kanshin et al. [33] retrospectively analyzed patients with cervical SCI. The authors found that the absence of F-wave during tibial nerve stimulation is a pathognomonic sign of spinal shock while restored registration determines ending of spinal shock. There are 4 consecutive clinical phases in SS (areflexia, initial recovery of reflexes, initial and constant hyperreflexia) [34]. Duration of each of these phases can vary depending on severity of SCI and anatomical spinal interruption that reduces accuracy of outcome prediction with ASIA impairment scale. F-wave registration is justified to confirm the completion of the first and the second phases of SS. Accurate clinical evaluation is possible after restoration of F-wave registration. Instrumental diagnosis of motor and sensory conduction by the methods of motor and somatosensory evoked potentials is advisable if ENMG is unavailable.

Application of neurophysiological survey for prediction of neurological outcome in victims with complicated SCI is shown in Table 1