Although the fundamental problems of upper motor neuron dysfunctions are neurological, the subsequent treatment and rehabilitation require an interdisciplinary approach. In the past years our institutions have conducted clinical as well as neuroscientific research in motor control. The Vienna Program for Movement Recovery (VPMR) integrates the efforts of these institutions into a joint research program with a common goal, to understand human neural control of movement and its recovery after severe neurological damage.
Epidural stimulation of the cervical and thoracic spinal cord for the control of chronic pain is the most common modality of all neuromodulation therapies. The lumbar spinal cord was targeted when epidural spinal cord stimulation (SCS) was applied for control of spinal spasticity of the lower limbs. In 1998 it was demonstrated that sustained, unpatterned lumbar epidural spinal cord stimulation can generate involuntary, automatic rhythmic movements of the paralyzed lower limbs in individuals with motor complete spinal cord injury (Rosenfeld et al. 1995, Dimitrijevic et al. 1998). Since the rhythmic motor outputs were generated in response to tonic input without manual manipulation of the lower limbs and in the absence of sensory feedback critical for rhythm generation (subjects were in supine position), evidence was provided for the existence of a pattern generator in the human lumbar spinal cord. This work was partially done in Vienna at the Maria Theresien Schloessel Neurological Hospital with Prof. Binder being the head of the institution. This demonstration and the question about the underlying mechanisms initiated the long lasting collaboration between the Viennese institutions under the mentorship of Prof. Milan R. Dimitrijevic (Baylor College of Medicine, Houston, TX; Foundation for Movement Recovery, Oslo, Norway). The first results of this cooperation demonstrated on a basis of electrophysiological and computational studies (first involvement of the Vienna University of Technology) that epidural lumbar spinal cord stimulation electrically directly activates afferent fibers but not directly interneurons and motoneurons. Thus, the lumbar neural circuits are activated transsynaptically through the afferent projections. This lead to the insight that the pattern and profile rather than the origin of the neural signals provided played an essential role in functionally activating the locomotor rhythm and pattern generators in the spinal cord. Specifically, the locomotor circuits were responsive to tonic input at a certain frequency range (20-60 Hz).
In the next generation of work the rhythm genesis was further elucidated (Minassian et al. 2004) and it was additionally shown that the human lumbar cord can generate the behavior of standing, the extension of the lower limbs with the controlled motor output to the lower limbs (Jilge et al. 2004). This was achieved by stimulating the same structures with the mere change of stimulation frequency (5-15 Hz). It should be also noted that by applying 50-100 Hz a strong suppression of motor outputs could be achieved, as essential for the control of spasticity (Pinter et al. 2000).
The expansion of the Viennese cooperation to an institute for physical medicine and rehabilitation motivated the development of non-invasive means of spinal cord stimulation. The activation of the target fibres with epidural leads depends mainly on their anatomical properties and less on the focus of the electrical field, thus specific stimulation of posterior roots is also feasible with a more distant, skin electrode based approach. The method of transcutaneous spinal cord stimulation, that activates at least a subset of the same neural structures as by epidural stimulation, was developed (Dimitrijevic et al. 2004, Minassian et al. 2007). Computer simulations of transcutaneous spinal cord stimulation confirmed that these anatomical factors are indeed essentially contributing to the low thresholds for posterior root stimulation (Ladenbauer et al., 2010; Danner et al., 2011). Amongst various potential applications of transcutaneous spinal cord stimulation, the elicitation of spinal reflexes for neurophysiological studies and the neuromodulative effect in applying continuous stimulation were the focus of subsequent studies.
I) Human neurosciences research programs
Theory of motor control and computational neurosciences
Volitionally induced motor-outputs recorded from the muscles display a so-called interference pattern reflecting various overlapping frequency components. Locomotor-like EMG activities induced by spinal cord stimulation, on the other hand, were composed of series of compound muscle action potentials (CMAPs) that can be unequivocally related to stimulus that triggered them (Minassian et al., 2004). These single responses within the locomotor-like EMG activities induced by spinal cord stimulation were identified as posterior root-muscle reflexes (PRM reflexes)—responses elicited in large-diameter posterior root afferents and detected from the muscles to which the responses were directed (Minassian et al., 2004; 2007). In response to constant stimulation the series of PRM reflexes were subject to rhythmic modulation processes.
This led us to hypothesize that the repetitive, electrically induced, inputs via multiple posterior roots evokes posterior root-muscle reflexes and concomitantly activated (due to the tonic signal profile) lumbar locomotor circuits via collateral branchings of the stimulated afferents. When set into action, the locomotor networks in turn modify the posterior root-muscle reflex activity. Thus, the variation in CMAP shape, amplitude and latency of successively elicited posterior root-muscle reflexes reflects the influence of interneuronal systems with intrinsic locomotor capacities and allow us to hypothesize about models of the human locomotor circuits.
Based on the elctrophysiological data, various models are formulated and tested using computational methods. Populations of neurons are modeled using Hodgkin-Huxley-like membrane equations and their dynamics in the proposed network models are observed to test and generate new hypotheses.
The posterior root-muscle reflex
The posterior root-muscle reflex of the triceps surae muscle group has some electrophysiological similarities with the soleus H-reflex, normally elicited by the stimulation of the posterior tibial nerve in the popliteal fossa. It has the following electrophysiological characteristics: (i) short and constant onset latencies of CMAPs; (ii) characteristic post stimulation depression of the CMAP amplitude; (iii) attenuation of CMAP amplitudes during Achilles tendon vibration; and (iv) increased amplitudes during slight plantar flexion and suppressed during voluntary contraction of the antagonistic tibialis anterior, with characteristic modifications of responses also in the other muscles. Owing to the convergence of all sensory fibres of the lower limbs close to the spinal cord, root stimulation can generate reflex responses in many muscles simultaneously and bilaterally. Posterior root-muscle reflexes can thus extend H-reflex studies of a single muscle to the simultaneous assessment of synaptic transmission of two neuron reflex arcs at multiple segmental levels. This is relevant because motor control involves many muscles with state and phase dependent changes of their functional roles, independently as well as within different synergies (Hofstoetter et al. 2008). Currently we are applying conditioning-test paradigms using the posterior root-muscle reflex to get a better insight in neurocontrol of human posture and locomotion.
Due the more central site of stimulation, the potential of mutual influence of reflexes through the multisegmental input as well as the pure sensory stimulation there can be also some differences between the behaviors of the posterior root-muscle reflex the H-reflex during conditioning-test paradigms. This requires a meticulous electrophysiological characterization of the posterior root-muscle reflex, along with a comparison with the H-reflex. We currently conducting several studies to compare well-described characteristic parameters and behaviors of the H-reflex with the ones of the posterior root-muscle reflex. These studies also involve our international collaborations with Ljubljana, Slovenia and Atlanta, GA.
II) Studies of novel restorative neurology interventions in motor disorders
Restorative neurology is a branch of neurology dedicated to improve functions of the impaired nervous system through selective structural or functional modification of abnormal neurocontrol according to underlying mechanisms and clinically unrecognized residual functions. The concept of restorative neurology to augment surviving central nervous system (CNS) capabilities to improve function is not new, yet it received renewed appreciation by the accumulated evidence for the existence of functional locomotor circuits in the human spinal cord that can execute complex stereotyped motor tasks in response to rather unspecific stimuli. These circuits are natural targets for new rehabilitation strategies since they likely remain intact even after severe upper motor neuron damage.
Rosenfeld, J. E., Sherwood, A. M., Halter, J. A. and Dimitrijevic, M. R. (1995). Evidence of a pattern generator in paralyzed subject with spinal cord stimulation. Society for Neuroscience Abstracts 21, 688. Washington, DC: Society for Neuroscience.