Transcutaneous spinal cord stimulation combined with locomotor training to improve walking ability in people with chronic spinal cord injury: study protocol for an international multi-centred double-blinded randomised sham-controlled trial (eWALK)

An international multi-centred, double-blinded, randomised sham-controlled trial (eWALK).

To determine the effect of 12 weeks of transcutaneous spinal stimulation (TSS) combined with locomotor training on walking ability in people with spinal cord injury (SCI).

Dedicated SCI research centres in Australia, Spain, USA and Scotland.

Fifty community-dwelling individuals with chronic SCI will be recruited. Participants will be eligible if they have bilateral motor levels between T1 and T11, a reproducible lower limb

muscle contraction in at least one muscle group, and a Walking Index for SCI II (WISCI II) between 1 and 6. Eligible participants will be randomised to one of two groups, either the active

stimulation group or the sham stimulation group. Participants allocated to the stimulation group will receive TSS combined with locomotor training for three 30-min sessions a week for 12

weeks. The locomotor sessions will include walking on a treadmill and overground. Participants allocated to the sham stimulation group will receive the same locomotor training combined with

sham stimulation. The primary outcome will be walking ability with stimulation using the WISCI II. Secondary outcomes will record sensation, strength, spasticity, bowel function and quality

of life.

Regaining the ability to walk is a priority for individuals with spinal cord injury (SCI) [1,2,3] and has been the focus of many studies and clinical trials [4]. While it is currently not

possible to restore voluntary control of muscles paralysed after SCI, recent developments have been made [5]. A series of case studies indicate that epidural spinal stimulation in

individuals with SCI can elicit step-like, rhythmic movement of the legs [6,7,8,9,10]. Similar effects have also been observed when transcutaneous spinal stimulation (TSS) is applied in

able-bodied individuals [11, 12], with evidence that TSS activates the same neural structures as epidural stimulation [13]. Moreover, there is preliminary evidence that TSS may be able to

restore voluntary movement and the ability to stand and walk in some individuals with SCI [14]. These improvements occur almost immediately in some individuals while others experience

progressive improvements after combined spinal stimulation and intensive physiotherapy. Moreover, there is preliminary evidence that TSS can reduce spasticity [15], which may also contribute

to improved walking ability.

Following a complete SCI, the human lumbosacral neural circuitry can generate rhythmic motor output in response to cutaneous and proprioceptive afferent signals elicited during assisted

standing and walking [14, 16,17,18,19,20]. Similarly, TSS can elicit rhythmic muscle activity [10, 21,22,23]. Importantly, these therapeutic modalities—assisted standing and walking and

TSS—interact with one another [24] and their combined application can elicit greater rhythmic muscle activity than either intervention alone [14, 22]. Moreover, repeated exposure to this

combined therapy can augment, and potentially restore, connections with supraspinal centres and postural reflex pathways [25,26,27,28,29]. As a population on the rise [30,31,32], individuals

with incomplete SCI are the most likely to regain the ability to walk [33], and TSS could help with recovery by increasing the excitability of lumbar spinal locomotor circuits that have

partially lost their descending motor drive [34,35,36,37]. Preliminary evidence indicates that TSS can improve gait kinematics and locomotor muscle activity in these individuals.

Despite positive preliminary results, a lack of evidence from randomised controlled trials has prevented translation into standard clinical practice [38]. Therefore, the primary aim of this

study is to determine the effectiveness of 12 weeks of TSS combined with locomotor training on walking ability in people with chronic SCI. For this purpose, we will compare a group

randomised to receive TSS plus locomotor training to a group randomised to receive sham TSS plus locomotor training. This study will also investigate the effect of this training on sensory

[39, 40], motor [41, 42] and bowel function [40, 43, see also 44], spasticity and quality of life. We hypothesise that TSS with locomotor training will improve walking ability in people with

chronic SCI who have some residual lower limb motor function. We also hypothesise TSS with locomotor training will improve spasticity, sensation, lower limb muscle strength, bowel function

and quality of life. Furthermore, we hypothesise that improvements will only be apparent in the presence of TSS.

An international multi-centred, double-blinded, randomised sham-controlled trial will be conducted in four SCI research centres located in Australia, USA, Scotland and Spain. Fifty

participants with chronic SCI will be randomised to one of two groups, namely the stimulation group or the sham group. Participants allocated to the stimulation group will receive TSS while

they complete 30-min of locomotor training, three times per week for 12 weeks (Fig. 1A). Participants allocated to the sham group will receive the same locomotor training, but with sham TSS.

Participant timeline (A) and schedule for capturing outcome measures (B).

Participants will be recruited via existing research databases, health professionals and advertisements on websites of SCI organisations. Potential participants will be given a copy of the

participant information sheet to read. They will be informed that their participation in the trial is voluntary and will not affect their current or future relationships with the study

centres. Participants will be encouraged to ask questions and to discuss the trial with family and friends before providing consent. Once essential trial information has been provided,

participants will be asked to give informed consent to participate in the trial by signing the Consent Form in the presence of a witness. These forms will be dated and retained by the

investigator at the site where consent is gained, and a copy will be provided to the participant. The present plan is for the Sydney site, the sponsor, to recruit approximately 20

participants and each of the other sites to recruit approximately 10 participants each.

A secure blocked random-allocation schedule will be computer-generated prior to the start of the trial by an independent person not directly involved in the trial. The random-allocation

schedule will consist of 50 three-digit codes: 25 assigned to the stimulation group and 25 assigned to the sham group. The three-digit codes will be used to program the stimulator control

unit. The random-allocation schedule will be uploaded to REDCap [45] to allow each study site to randomise participants. To maintain blinded allocation, the REDCap random-allocation schedule

will display the three-digit code, not the participant’s allocated group. Once a participant is deemed eligible for the trial, a baseline assessment will be completed, and a study

investigator will randomise the participant within REDCap. The participant will be considered enroled in the trial at this time.

The physiotherapists and assistants delivering the locomotor training will be blinded to treatment allocation. Participants will be instructed to not discuss their perceived group allocation

with the physiotherapists and assistants. The success of blinding will be recorded at week 12. When enroled in the trial, participants will be asked to stop any formal physiotherapy

programs aimed at improving walking.

Both groups will receive three 30-min locomotor training sessions per week for 12 weeks while receiving their allocated stimulation. Previous studies examining transcutaneous stimulation and

locomotor training protocols vary considerably in frequency and duration [46], particularly stimulation studies in which there are currently no RCTs [14, 41]. Many of the studies examining

the effects of traditional locomotor training programs, typically train participants between 8 to 12 weeks, 3–5 times a week, with improvements in walking ability seen in people with

incomplete SCI [47,48,49]. Potential participants were contacted to gauge whether it would be feasible for them to train three times a week for 12 weeks, this was deemed feasible by all the

people contacted. Therefore, three 30-min locomotor training sessions per week for 12 weeks was chosen as an adequate training dose, taking into account current established training

protocols and the time commitment from participants. While overground and treadmill locomotor training have equivalent therapeutic benefits in people with incomplete SCI [24,25,26],

body-weight support treadmill training is more ergonomic for therapists and allows for greater dosage (i.e., steps) when training individuals with impaired mobility (i.e., WISCI II scores

from 1 to 6). Thus, participants will undergo five treadmill and one overground training session(s) per fortnight, for a total of 30 treadmill and six overground training sessions over 12

weeks. Training will be delivered by an experienced neurological physiotherapist and up to two trained assistants as required. Training will be provided to all therapists at each site on how

to perform the locomotor training to ensure consistency across sites.

Participants will be allowed standing, and if needed, seated rests during the training sessions. Stimulation will be stopped during seated rests, and this time will not count towards the 30 

minutes training target. In contrast, stimulation will continue during standing rests, and this time will count towards the 30 minutes training target. For each session, a maximum of 60 

minutes will be allotted to complete the 30 minutes of training. While it is possible a participant may not complete 30 minutes of training in a session if they require regular or prolonged

seated rests, unpublished pilot work conducted in the Sydney centre found that participants rarely require seated breaks. Standing breaks, when taken, typically lasted no longer than 1–2 

minutes. Thus, we expect most participants to complete the 30 minutes of training each session, with > 85% of each session spent walking.

Manual assistance of the lower limbs based on established locomotor protocols will be provided as required to improve walking patterns [50]. The amount of body-weight support required will

be assessed on the first training day. Excessive knee flexion during the stance phase (i.e., > 40°) or toe dragging during the swing phase will indicate body-weight support needs to be

increased [51]. Body-weight support will be reviewed regularly with the aim of walking with the greatest amount of weight-bearing that does not cause excessive knee flexion or toe drag.

There will be no limit to the amount of body-weight support that can be provided, as long as the participant’s heels contact the treadmill. The speed of the treadmill will be monitored and

adjusted by the physiotherapist throughout the 12-week training period, with the aim of walking at the fastest comfortable speed that allows for a correct walking pattern. For overground

training sessions, participants will be encouraged to walk at a comfortable pace. To ensure safety and optimise the participant’s walking pattern during overground training sessions,

orthoses, gait aides, parallel bars and safety harnesses will be used as required.

The details of each training session will be recorded by the therapist in a training diary. Changes in pain and spasticity will be recorded on a weekly basis. Changes in medication use will

also be recorded.

Transcutaneous spinal stimulation will be applied with the anode (5 × 10 cm) placed over the lower abdomen and the cathode (5 × 10 cm) placed over the lower back, both centred on the midline

of the body [14, 22, 52, 53]. For the anode, the long edge of the electrode will be oriented horizontally. For the cathode, the long edge of the electrode will be oriented vertically. The

preferred cathode placement will be with the top edge of the electrode in-line with the L1–L2 vertebral interspace. However, this location may be adjusted rostrally—to a maximum of the

T11–T12 vertebral interspace—to accommodate for implanted metal hardware or the rare case that posterior root-muscle (PRM) reflexes cannot be elicited at the preferred electrode site. Prior

to the first day of locomotor training plus stimulation (or sham), all participants will be briefed on what to expect with regards to the stimulation. Specifically, participants will be

informed that each person experiences the stimulation differently based on their level of sensation and tolerance. Participants will also be informed that the sensations associated with the

stimulation may change over time, including over the course of a single training session.

Studies, case studies and case series investigating the use of TSS to improve walking ability in people with SCI have used a variety of stimulation parameters [54]. The choice of parameters

for the current trial was based on a study conducted in Sydney on 10 able-bodied participants and 10 participants with SCI [55] and a survey of the literature. The basic stimulation waveform

will be 1 ms in duration, filled with a biphasic 10 KHz carrier frequency. This type of waveform has been used in several previous studies [12, 14, 39, 56,57,58]. While some studies have

reported similar effects without the high carrier frequency [22, 41, 53], there is evidence that, with the high-frequency component present, motor responses can be elicited with less

discomfort [59, 60], although a recent paper does not support this claim [61]. Moreover, there is recent evidence that, for the upper limb, including a carrier frequency may have suppressive

effects on cortical excitability in people with SCI, which is associated with greater functional performance [62]. However, the mechanisms underlying improvements remain unclear for

conventional or carrier frequency stimulation delivered either epidurally or transcutaneously.

Posterior root-muscle reflexes are evoked muscle responses elicited by electrically stimulating the posterior nerve roots of the spinal cord. To assess these, biphasic rectangular single

pulses of 1 ms with a 10 kHz carrier frequency will be delivered using a Digitimer Biphasic Constant Current multi-modal stimulator (DS8R, Digitimer Ltd, UK), driven by a custom stimulator

control unit. The minimal stimulation intensity required to induce PRM reflexes in the bilateral vastus medialis muscles (> 50 μV peak-to-peak amplitude above the background muscle activity

in 5 of 10 consecutive trials in relaxed muscles) will be used to set TSS intensity during locomotor training [62]. Reflexes will be assessed in standing with approximately 90% body-weight

support to enable some afferent feedback from the feet, yet no EMG activity from the vastus medialis muscles. The location of the EMG electrodes will be standardised across sites. This

threshold will be assessed prior to randomisation and then re-assessed every 2 weeks, during the training period.

Transcutaneous spinal stimulation during locomotor training will consist of the same biphasic rectangular pulse delivered tonically at 20 Hz. TSS intensity will be 100% of the threshold

intensity determined during the most recent PRM reflex testing session. Various TSS intensities have been used in studies involving people with SCI, with beneficial effects reported for

intensities that are below the motor threshold, above the motor threshold and at the motor threshold [41, 62, 63]. We chose to not use a subthreshold target intensity as this is difficult to

set reliably, especially across therapists, sessions and study sites [64]. Moreover, a study conducted in Sydney confirmed that the proposed TSS intensity elicits only very small lower limb

muscle contractions, which do not interfere with the locomotor training or with walking by the participants.

Various stimulation frequencies have been used to induce locomotor-like muscle activity using TSS, ranging from 5–50 Hz [14, 22, 40, 65]. The amplitude of step-like movements increase as the

stimulation frequency increases from 5 Hz to 40 Hz [12]. More recently, lower stimulation frequencies ( 90% probability of detecting a between-group difference of 2 points on the primary

outcome: WISCI II [73]. This assumes an alpha of 0.05, a SD of 2 points [48, 78] and a dropout rate of 15%.

We will use an intention-to-treat analysis to draw accurate and unbiased conclusions regarding the effectiveness of our intervention. That is, participants will be analysed according to the

group to which they were allocated, regardless of compliance with the intervention. Statistical analysis will be conducted blind to treatment allocation.

All outcomes will be analysed with multi-level (i.e., mixed) models. Ordinal measures with few scale values (e.g., Modified Ashworth Scale) will be dichotomised and assessed using

multi-level logistic regression. Continuous measures and ordinal measures with many scale values (e.g., WISCI II) will be assessed using multi-level linear regression. In both cases, the

participant will be a random factor with a random intercept. We will verify the appropriateness of statistical procedures (diagnostic tests) a priori and, if required, identify appropriate

alternative analyses (e.g., data transformations, robust analyses) and the order in which they should be applied.

We are primarily interested in whether improvements in outcomes differ between the stimulation group and the sham group following 12 weeks of locomotor training. Baseline values will be

included as a covariate. All contrasts will be performed and results reported as mean effects and 95% confidence intervals.

All adverse events will be recorded and reported to the Principal Investigator. The Principal Investigator will be responsible for reporting any serious adverse events to the Ethics

Committee as soon as possible. All adverse events and serious adverse events will be followed until they have abated, or until a stable situation has been reached. Depending on the event,

additional tests or medical procedures may be required, as well as a review by a general medical practitioner or SCI physician.

All information collected for this trial will be de-identified and kept confidential and secure. All files containing participants’ personal details will remain at the trial site where they

were collected. Moreover, case report forms will only contain participant ID codes and upon trial completion will be stored at the trial site where they were collected. Electronically

transcribed data will be stored on the secure REDCap system managed by Neuroscience Research Australia. Access to data will only be granted to the Principal Investigators and other research

staff directly involved in the study. Individual names of the participants will not be considered in data analysis and participants will not be identified in published data. Any data stored

for future analysis will be de-identified.

Trial monitoring will be undertaken by the Principal Investigator, an independent Data Monitoring and Safety Committee (DMSC), and an independent trial monitor. Best practice conduct of the

trial will be ensured through frequent monitoring by the responsible Investigators and the clinical trial monitor, with the purpose of facilitating the work and fulfilling the objectives of

the trial.

All design features important for minimising bias will be adhered to and the trial has been registered with the Australian and New Zealand Clinical Trials register (ACTRN12620001241921).

This study has been approved by the ethics committee in Sydney and is currently awaiting approval in the other sites. We certify that all applicable institutional and governmental

regulations concerning the ethical use of human volunteers will be followed during the course of this research.

Results will be presented at national and international conferences or similar. Participant’s individual results will be available on request from the Principal Investigator at their site.

Data sharing is not applicable to this article as no datasets were generated or analysed. For the main trial, full de-identified data used to generate all results will be made available with

the publication.

The authors acknowledge Professor Rob Herbert/Peter Humburg for their statistical advice; SpinalCure Australia, Spinal Cord Injuries Australia and Paraquad for their assistance with an

advertisement for this trial; and the support of local Spinal Cord Injury Units at each site.

Funding for this study has been received from SpinalCure Australia and Catwalk NZ.

Neuroscience Research Australia, Randwick, NSW, 2031, Australia

Elizabeth A. Bye, Martin E. Héroux, Claire L. Boswell-Ruys, Bonsan B. Lee, Euan J. McCaughey, Jane E. Butler & Simon C. Gandevia

School of Medical Sciences, University of New South Wales, Kensington, NSW, 2052, Australia

Elizabeth A. Bye, Martin E. Héroux, Claire L. Boswell-Ruys, Bonsan B. Lee, Euan J. McCaughey, Jane E. Butler & Simon C. Gandevia

Shirley Ryan Ability Lab, Northwestern University, Hine VA Hospital, Chicago, USA

Queen Elizabeth National Spinal Injuries Unit, Queen Elizabeth University Hospital, Glasgow, G51 4TF, Scotland

Hospital Nacional de Parapléjicos, SESCAM, Toledo, 45071, Spain

Harris Manchester College, University of Oxford, Oxford, OX1 3TD, UK

Conceptualisation: EAB, MEH, CLB, BBL, EJM, JEB, and SCG; Methodology: EAB, MEH, CLB, BBL, EJM, JEB, and SCG; Writing—original draft: EAB, EJM, and MEH; Writing—review & editing: All

authors; Project administration: EAB, MEH, CLB, BBL, EJM, JEB, and SCG; Funding acquisition: SG and JB.

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Anyone you share the following link with will be able to read this content: