Non-Invasive Neuromodulation and Rehabilitation to Promote Functional Restoration In Persons with Spinal Cord Injury

Structured Abstract

Purpose of review:

This review will focus on the use of clinically accessible neuromodulatory approaches for functional restoration in persons with spinal cord injury (SCI).

Recent findings:

Functional restoration is a primary rehabilitation priority for individuals with SCI. High-tech neuromodulatory modalities have been used in laboratory settings to improve hand and walking function as well as to reduce spasticity and pain in persons with SCI. However, the cost, limited accessibility, and required expertise are prohibitive for clinical applicability of these high-tech modalities. Recent literature indicates that non-invasive and clinically accessible approaches targeting supraspinal, spinal, and peripheral neural structures can modulate neural excitability. Although a limited number of studies have examined the use of these approaches for functional restoration and amelioration of secondary complications in SCI, early evidence investigating their efficacy when combined with training is encouraging.

Summary:

Larger sample studies addressing both biomarker identification and dosing are crucial next steps in the field of neurorehabilitation research before novel non-invasive stimulation approaches can be incorporated into standard clinical practice.

Introduction

Restoration of upper and lower extremity function in persons with spinal cord injury (PwSCI) share several principles in common. Neural circuits at all levels of the neuraxis demonstrate use-dependent plasticity in response to repeated activation, wherein decreased use results in decrement in capacity, and increased use is associated with improved efficacy of those circuits.(1–3) This plasticity is the basis for all training-related changes in performance; and in fact, training, practice, and other forms of repetitive activities intended to improve cognitive and behavioral functioning are the original “neuromodulators”. As such, whether directed at improving motor function in the upper extremities or lower extremities, neuromodulation approaches that are intended to promote adaptive plasticity target the same neural mechanisms that are targeted through practice and training.(4–6)

Beyond the common mechanisms of use-dependent plasticity, control of both the upper and lower extremities depends to a large extent on descending supraspinal drive. Connections between supraspinal circuits and the motoneurons controlling hand function are among the most direct in the nervous system,(7, 8) and for this reason damage to descending tracts has an obvious effect on hand function. Supraspinal influences on lower extremity function are also important; this descending input is essential for step initiation, activation of spinal central pattern generators, and other aspects of locomotion.(9–11) After SCI, damage to spinal tracts disrupts information transmission from the brain to the spinal cord, and results in maladaptive reorganization of the entire neuraxis that contributes to motor impairment.(12, 13) For example, in addition to the damage caused by the injury itself after SCI, maladaptive cortical plasticity and reductions in cortical excitability (14) further reduce descending supraspinal drive. At the level of the spinal cord, maladaptive reorganization of spinal circuits results in spasticity, which is due in part to loss of descending control of inhibitory spinal circuits.(15)

From a motor control perspective, the maladaptive changes to supraspinal circuits described above means that after SCI the brain is less effective than it could be at driving information through the residual spinal pathways. Training combined with clinically accessible forms of stimulation directed at increasing corticospinal excitability is associated with functional improvements in PwSCI.(16–18) Therefore, stimulation modalities that target cortical excitability, such as non-invasive brain stimulation (NIBS), or indirectly target cortical circuits using peripheral nerve stimulation, is an area of great interest in SCI neurorehabilitation research. In addition to being clinically accessible, these forms of NIBS have relatively minor adverse effects.(19) Itching is the most common complaint.(20) Use of stimulation is always associated with a risk of skin irritation and in extreme cases with blistering, and some individuals report post-stimulation headaches.

In addition to the use of stimulation to activate cortical circuits, stimulation can also be used to engage spinal circuits. Stimulation of peripheral nerves and spinal nerve roots preferentially activates the large-diameter afferent fibers,(21–23). Afferent input represents a robust approach to activating spinal circuits, with effects that can increase or decrease motoneuron and reflex excitability.(24–27) As such, afferent input can be a valuable adjunct to rehabilitation interventions for restoration of both hand function and walking function.

Although many stimulation modalities have been successfully implemented in the research setting, their application as a part of standard neurorehabilitation or home training programs is limited due to lack of clinical accessibility. Compared to high-tech NIBS modalities, transcranial current stimulation (tCS) modalities are less costly (28–30), easier to apply (30, 31), and do not require high-tech equipment (29) or immobilization of the head.(32) In regard to afferent stimulation, transcutaneous spinal stimulation (TSS) is a non-invasive alternative to epidural spinal stimulation that can be applied via surface electrodes.(33, 34) Importantly, TSS has been shown to activate the same spinal circuity as epidural stimulation.(21, 35) As with other forms of stimulation, TSS can be associated with skin irritation and in extreme cases with blistering. Simulation can be uncomfortable at high intensities. This review will focus on studies of clinically accessible stimulation modalities in PwSCI.

Neuromodulation and rehabilitation in the restoration of hand function

There are a limited number of studies investigating the use of transcranial direct current stimulation (tDCS) for upper extremity functional restoration in PwSCI, and these have yielded divergent results. Although some studies demonstrate that tDCS is associated with beneficial effects, including increased corticospinal excitability (36), grasp movement efficiency (37), and functional performance (38), other studies report that tDCS has no effects.(39, 40) Inter-individual variability in tDCS responsiveness (41–45) likely contributes to the reported heterogeneity in tDCS efficacy. However, because inter-individual variability has only been reported in non-injured individuals, we will focus on the potential contributions of variations in tDCS dosing (i.e. intensity, electrode size, and stimulation montage) for this review.

Regarding tDCS dosing, tDCS intensities of 1 and 2 mA and durations of 20 and 30 minutes have been investigated in PwSCI. Although single-session application of 1 mA tDCS is associated with a small effect on corticospinal excitability (38), data from subsequent randomized crossover studies suggest that 2 mA stimulation is the minimum tDCS stimulation intensity needed to induce changes in corticospinal excitability (36) and enhance motor performance (37) in PwSCI. The effects of single-session application of 2 mA tDCS on corticospinal excitability do not persist 20 minutes post-stimulation, indicating that the duration of tDCS after-effects for upper extremity muscles may be shorter than those reported in non-injured individuals.(46, 47) While the transience of single-session effects of tDCS in PwSCI suggests that multiple tDCS sessions may be needed to induce persistent effects, the two studies investigating multi-session application of tDCS for upper extremity function have yielded inconclusive results.(39, 40)

In addition to stimulation intensity, the efficacy of tDCS is also dependent upon electrode size and montage.(48–50) A range of tDCS electrode sizes have been utilized in PwSCI (3.14–35 cm²). While tDCS focality can be improved by using smaller electrodes (51, 52), recent in vitro research proposes that functional targeting via the pairing of tDCS with training is a primary determinant of the focality of tDCS-induced after-effects.(4) Considerably less attention has been dedicated to electrode montage in PwSCI. Although the use of conventional unihemispheric electrode montages produce less intra-individual variability when compared to high-definition montages (53), unihemispheric stimulation may not be the most efficacious approach for improving upper extremity function in PwSCI. Because upper extremity functional impairments after SCI are typically bilateral, bihemispheric excitatory stimulation intended to target the motor cortex representations of both extremities concurrently may be a more suitable for this population. However, it has been shown that two excitatory electrode montages are required to increase bimanual hand performance in non-injured individuals.(54–56) We have shown that the application of bihemispheric excitatory tDCS is feasible in individuals with tetraplegia.(57)

As tCS research continues to expand, the efficacy of patterned tCS modalities, including transcranial pulsed current stimulation (tPCS) and transcranial random noise stimulation (tRNS), is being investigated. Studies have demonstrated that patterned stimulation has a more robust influence on neural excitability than uniform stimulation (58, 59), suggesting that tCS modalities with patterned waveforms may be more efficacious modulators of corticospinal excitability than the constant current delivered with tDCS. Comparative studies of tPCS and tDCS in non-injured individuals have demonstrated that tPCS is associated with larger increases in corticospinal excitability of upper extremity muscles than tDCS.(60, 61) Similarly, tRNS has been shown to induce more consistent (62) and larger (62, 63) cortical excitatory effects for upper extremity muscles when compared to tDCS. Both tPCS and tRNS are well tolerated (57, 64) when combined with upper extremity functional task practice in PwSCI. Further, tRNS improves the cortical excitability of upper extremity muscles as indicated by a moderate effect size.(64) This encouraging preliminary evidence regarding the efficacy of patterned tCS modalities in PwSCI warrants future research.

Both electrical stimulation and focal vibration have been investigated as afferent stimulation modalities for improving upper extremity function. Somatosensory stimulation of the median nerve improves upper extremity strength (65), function (16, 65), and sensation (66) when combined with practice. Enhanced hand function has also been observed with use of upper extremity functional electrical stimulation (FES) (16) and transcutaneous electrical nerve stimulation (TENS).(38) In addition to changes in clinical outcomes, electrical stimulation is associated with enhanced corticospinal excitability as determined by reduced motor thresholds (66) and increased corticospinal map area/volume (16, 67) and motor evoked potential amplitude.(38) Similar to peripherical stimulation modalities, single-session application of cervical spinal stimulation is also associated with enhanced hand motor performance and increased cortical and spinal excitability in PwSCI.(68) Multi-session application of combined cervical TSS and training results in improvements in hand strength and function that persist for a minimum of 3 months following intervention conclusion.(69) Finally, afferent stimulation via focal vibration of the distal tendon of the flexor carpi radilais is associated with short term persistence (30 minutes) of enhanced pinch force and motor evoked potential amplitude.(38) Together, these data suggest that plasticity of both the spinal and corticospinal circuitry contribute to the effects of afferent stimulation on upper extremity function in PwSCI.

Neuromodulation and rehabilitation in the restoration of walking function and management of spasticity and pain

Walking

In PwSCI, locomotor training is commonly used with the goal of improving walking function, (70) and this approach is grounded in the literature related to use-dependent plasticity.(71) Questions related to the dose of training needed to elicit neuroplastic effects remain to be answered as a wide range of approaches and intervention periods have been studied.(72) Early evidence suggests intensity of training (73) and the environment in which training occurs (ie, overground versus treadmill) (74) have an influence on outcomes. However, identifying the individuals who are likely to obtain functional benefit from this training is important.(75)

The importance of supraspinal inputs for activating spinal circuits that contribute to walking makes tDCS a potentially valuable intervention. One advantage of tDCS targetting cortical regions associated wtih the lower extremities is that a single electrode montage can be used to target both lower extremities.(76) However, evidence for the value of tDCS to improve lower extremity function in PwSCI is limited. The two published studies investigating the multi-session application of tDCS for lower extremity function yielded varied results. While no effects of tDCS were reported after 20 sessions of tDCS combined with robot-assisted gait training (77), 36 sessions of tDCS were associated with unilateral increases in lower extremity manual muscle test scores.(78)

The use of peripheral nerve stimulation as an adjunct to locomotor training has a long history in the context of FES wherein the intent is direct muscle activation. While the use of stimulation expressly to promote neuromodulation is a newer purpose, it is noteworthy that FES itself has been shown to have neuromodulatory effects in PwSCI.(79, 80) TSS can be considered a form of peripheral nerve stimulation, as its primary effects are on large diameter afferents of the spinal nerve roots. Investigation of TSS as an adjunct to locomotor training is still in the early phases. There is some evidence that TSS can improve walking kinematics,(81) and appears to augment the effects of locomotor training in persons with motor-incomplete SCI.(82) Randomized clinical trials indicate that use-dependent plasticity augmented by neuromodulatory stimulation can influence the performance of motor behaviors such as walking, (74, 82) as well as modulate excitability associated with spasticity and pain.(24, 83)

Spasticity

Spasticity is a common and problematic secondary condition in PwSCI that is a manifestation of numerous aberrations of neural control. Among these are damage to descending tracts with associated loss or impairment of excitation to spinal presynaptic inhibitory circuits,(15) a shift in the chloride equilibrium potential resulting in decreased inhibitory influence of GABA,(84) and the advent of constitutively active serotonin receptors resulting in persistent inward currents that decrease the firing threshold of the motoneuron.(85) There is evidence that these impairments can be influenced by motor training and/or by stimulation.(80, 86–88)

PwSCI who experience spasticity report that movement-related activities are of greater value for managing spasticity than antispasmodic medications.(89) Both active and passive movements produce post-activation depression of spinal circuits.(90–92) Consistent with this idea, evidence indicates that locomotor training is associated with improved modulation of spinal reflexes and reduction in lower extremity spasticity.(80, 87, 93)

Since loss of supraspinal inputs to spinal inhibitory circuits is one of the neural origins of spasticity after SCI,(15) it is possible that increasing supraspinal drive through the remaining spinal pathways could improve activation of these inhibitory circuits. However, the only study investigating the use of tDCS for reduction of lower extremity spasticity in PwSCI concluded that single-session application of tDCS had no effect.(24)

As described earlier, afferent input in the form of stimulation has potent neuromodulatory effects on spinal circuits. TENS, which has long been used to modulate the excitability of spinal circuits associated with pain, appears to have value for the management of spasticity in PwSCI.(94) Likewise, early evidence suggests that TSS may also have value for spasticity management.(95)

Pain

Beyond spasticity, pain is another secondary complication of SCI related to aberrant neural reorganization that occurs along the neuraxis.(96) Therefore, utilizing brain stimulation for pain management may be a useful alternative approach to pharmacological treatments.(97) Reduction of pain has been reported in several studies in association with the application of tDCS over the motor cortex in PwSCI.(98, 99) When tDCS was combined with visual illusion techniques, its analgesic effects were augmented and lasted up to 12 weeks following the intervention.(100) Conversely, multiple sessions of tDCS did not ameliorate pain intensity in a group of individuals long-term chronic SCI.(101)

Afferent input has long been used for management of pain, a fact that we recognize instinctively when we immediately rub our elbow after having accidentally slammed it into the door. Activation of large-diameter afferents is a key feature of the gate control theory of pain modulation.(102) Epidural stimulation is a well-established approach for management of chronic pain(103); while once referred to as “dorsal column stimulation”, modeling studies indicate that stimulation effects likely arise from activation of large-diameter afferents in the spinal roots, which are accessble via TSS.(35, 104) Modulation of pain via the large-diameter afferents is the theoretical mechanism underlying the use of TENS, and this form of stimulation has been shown to be of value for management of pain (including neuropathic pain) in PwSCI.(105, 106)

Conclusions

In summary, clinically accessible modalities (including tCS, peripheral nerve stimulation and TSS) have great potential value as adjuncts to conventional therapeutic protocols due to the ease of application, cost effectiveness, and demonstrated neuromodulatory and functional effects. Thought leaders in rehabilitation research have indicated that combinatorial approaches are most likely to have meaningful impact.(107) A 2016 systematic review of all the SCI clinical trials intending to improve motor function published up to that time supports this perspective. The review found that the strongest evidence for meaningful outcomes was for studies wherein the experimental pharmacological or device intervention was combined with rehabilitation.(108) Recent evidence from preclinical models has identified the important role that movement plays in plasticity after SCI.(109) Further, the optimal time window for the application of interventions is likely the period early after injury as suggested by evidence from pre-clinical models of spinal cord injury [see Fouad 2011 (110) for review]. However, many studies of chronic SCI in both pre-clinical and human trials have shown that the potential for plasticity persists into the chronic phase of SCI.(111, 112)

Although current literature regarding the use of cortical and afferent stimulation in PwSCI is encouraging, studies with increased sample sizes are needed to address questions regarding variability in stimulation responsiveness and dosing before novel stimulation modalities can be integrated into standard clinical practice. In general, there is a large degree of variability in injury severity and the resulting functional deficits in PwSCI. This inherent population heterogeneity highlights the need for identification of robust, easily acquired biomarkers that can predict stimulation responsiveness. Because variability in responsiveness is a complex phenomenon, the most promising biomarkers will likely be derived from a combination of multiple outcomes (i.e., genetic, clinical, neurophysiologic). Research investigating optimal dosing parameters for stimulation, including intensity, duration, number of sessions, electrode size, and montage, is also needed. Finally, studies examining the comparative efficacy of different stimulation modalities will aid in the identification of the most promising novel stimulation modalities. Future randomized controlled trials investigating these three priorities will further advance the field of SCI neurorehabilitation research.

Key points:

• Neuromodulatory stimulation approaches activate many of the same circuits that are activated by training at all levels of the neuraxis, and therefore have the potential to reverse maladaptive reorganization that is associated with spinal cord injury.

• Clinically accessible neuromodulatory approaches targeting cortical and spinal circuits can augment the therapeutic effects of training for functional restoration in persons with spinal cord injury

• Future clinical trials should focus on investigating questions related to dosing, biomarkers of responsiveness, and personalizing protocols based on individual characteristics in persons with spinal cord injury.