enabling task-specific volitional motor functions via spinal cord … · 2017. 8. 8. · functions...

BRIEF REPORT

From the DepartmentNeurologic Surgery (P.JI.A.L., D.I.D., A.A.M., K.HRehabilitation MedicineResearch Center, Depament of Physical Medicand Rehabilitation (M.GM.L.G., J.A.S., L.A.B., M.BA.R.T., C.L., K.D.Z., K.HMayo Clinic Graduate Sof Biomedical Sciences(J.S.C.), and DepartmenPhysiology and BiomedEngineering (K.H.L.), MaClinic, Rochester, MN;Department of IntegratBiology and Physiology,versity of California LosAngeles (D.G.S., P.N.G.Y.P.G., V.R.E.); and PavlInstitute of Physiology,Russian Academy of SciSt. Petersburg, Russia (Y

544

Enabling Task-Specific Volitional MotorFunctions via Spinal Cord Neuromodulation in a

Human With Paraplegia

of.G.,.L.),

rt-ine.V.S.,.L.,.L.),chool

t oficalyo

iveUni-

,ov

ences,.P.G.).

Peter J. Grahn, PhD; Igor A. Lavrov, MD, PhD; Dimitry G. Sayenko, PhD;Meegan G. Van Straaten, PT; Megan L. Gill, DPT; Jeffrey A. Strommen, MD;

Jonathan S. Calvert, BS; Dina I. Drubach, DScPT; Lisa A. Beck, MS;Margaux B. Linde, BS; Andrew R. Thoreson, MS; Cesar Lopez, MS;

Aldo A. Mendez, MD; Parag N. Gad, PhD; Yury P. Gerasimenko, PhD;V. Reggie Edgerton, PhD; Kristin D. Zhao, PhD; and Kendall H. Lee, MD, PhD

Abstract

We report a case of chronic traumatic paraplegia in which epidural electrical stimulation (EES) of thelumbosacral spinal cord enabled (1) volitional control of task-specific muscle activity, (2) volitional controlof rhythmic muscle activity to produce steplike movements while side-lying, (3) independent standing,and (4) while in a vertical position with body weight partially supported, voluntary control of steplikemovements and rhythmic muscle activity. This is the first time that the application of EES enabled all ofthese tasks in the same patient within the first 2 weeks (8 stimulation sessions total) of EES therapy.

ª 2017 Mayo Foundation for Medical Education and Research n Mayo Clin Proc. 2017;92(4):544-554

I t is well accepted that severe spinal cordinjury (SCI) leads to functional disconnec-tion of ascending and descending spinal

pathways, impairing neural circuitry throughand below the SCI. However, in clinicallycomplete SCI, defined as American SpinalInjury Association Impairment Scale (AIS)level A, a portion of neural tissue commonlyremains intact across the injury site1 andmay provide nonspecific supraspinal influenceon sublesional spinal circuitry.2-5 This injuryprofile is defined as “discomplete.”

Anemerging therapy for recovering functionafter SCI,whichmaybemore effective in injuriesidentified as discomplete, is epidural electricalstimulation (EES). In patients with SCI diag-nosed as having AIS-A or AIS-B (in the latter,some sensory function remains intact), EES ofthe lumbosacral spinal cord has been shown todrive nonvolitional rhythmic motor circuitry.6-9

In a recent paramount study, EES facilitatedvolitional control of joint-specific muscles andindependent standing after months of trainingwith EES (Supplemental Table, available onlineat http://www.mayoclinicproceedings.org).10,11

Mayo Clin Proc. n April 2017www.mayoclinicproceedings.org n

Although these findings were powerful, theyhave yet to be reported by other research teams,a necessary milestone before these treatmentscan be widely adopted by the medical commu-nity. Therefore, the initial goal of this studywas to replicate the findings from work per-formed at the University of Louisville that re-ported that (1) EES enabled volitional controlof motor activity and (2) EES enabled indepen-dent standing.10,11 In addition to the initialgoal of replication, we set out to determinewhether EES could enable volitional controlover rhythmic, steplike activities.

METHODS

Participant DescriptionAll the procedures for this study were per-formed with the approval of the Mayo ClinicInstitutional Review Board and with an Inves-tigational Device Exemption from the USFood and Drug Administration. The partici-pant was a 26-year-old man who sustained atraumatic T6 AIS-A SCI 3 years before studyenrollment. Immediately after his injury, he

;92(4):544-554 n http://dx.doi.org/10.1016/j.mayocp.2017.02.014ª 2017 Mayo Foundation for Medical Education and Research

http://www.mayoclinicproceedings.org

http://crossmark.crossref.org/dialog/?doi=10.1016/j.mayocp.2017.02.014&domain=pdf

http://dx.doi.org/10.1016/j.mayocp.2017.02.014


SPINAL CORD NEUROMODULATION AFTER PARALYSIS

underwent a spinal fusion of the fifth toeleventh thoracic vertebrae and participatedin in-patient physical rehabilitation forapproximately 2 months to achieve indepen-dence during activities of daily living from awheelchair. Once discharged from in-patientrehabilitation, the patient underwent 5 weeksof outpatient SCI rehabilitation. After those 5weeks, he performed only self-directed upperlimb strengthening and general lower limbstretching without any other form of formalneuromuscular training until enrollment inthis study.

Study TimelineOn enrollment, the patient’s motor andsensory characteristics were documented byclinical examination and electrophysiologicmeasures. The American Spinal Injury Associ-ation Examination as defined by the Interna-tional Standards for the Classification ofSpinal Cord Injury was used to determinethe extent of the SCI, including the neuro-logic level and the “completeness” of theinjury. Electrophysiologic measures includedupper and lower limb somatosensory evokedpotentials and transcranial magnetic motorevoked potentials recorded over select upperand lower limb muscles. This was followedby 22 weeks of motor training. An EESsystem was then implanted in the region ofthe lumbar enlargement, followed by 3 weeksof postoperative recovery and 8 sessions ofvolitional motor performance testing in thepresence of EES over a 2-week period(Figure 1A).

Electrophysiologic Assessment ofTranslesional Neural ConnectivityTo test for potentially intact translesional neu-ral connectivity below the threshold for stan-dard clinical electrophysiologic recordings,we applied separate conditioning stimulibefore spinally evoked motor potentials eli-cited via transcutaneous stimulation over theregion of the lumbar spinal cord enlarge-ment.12-21 These evoked potentials wererecorded over select lower limb muscles. Theconditioning stimuli were applied in anattempt to activate (1) long propriospinaltracts via ulnar nerve stimulation, (2) descend-ing white matter via cervical spinal cord stim-ulation, or (3) the corticospinal tract via

Mayo Clin Proc. n April 2017;92(4):544-554 n http://dx.doi.org/10.1www.mayoclinicproceedings.org

transcranial magnetic stimulation of the motorcortex. The conditioning stimuli were fol-lowed 5 to 200 milliseconds later by test stim-uli delivered to the lumbosacral spinal cord toelicit spinally evoked motor potentials inlower limb muscles.

Motor Training ParadigmBefore implantation of the EES system, thepatient underwent 61 motor training sessions(approximately 3 sessions per week for 22weeks) performed by a team led by a phy-sical therapist (M.L.G.) and a kinesiologist(M.B.L.), both of whom had extensive loco-motor education from the NeuroRecoveryTraining Institute (http://www.neurorti.com).22 Sessions consisted of approximately15 minutes of lower extremity stretching toensure optimal kinematics, 45 minutes of lo-comotor training on a treadmill with bodyweight support and trainer assistance at thelegs and pelvis, and 30 minutes of balanceand task-specific strengthening exercisesduring sitting and while standing incustom-built standing bars with trainerassistance at the knees and pelvis(Supplemental Figure 1A-D, available onlineat http://www.mayoclinicproceedings.org).During select sessions at approximately4-week intervals, the patient was instructedto first attempt whole-leg, then knee, andthen ankle flexion/extension of each legwhile side-lying with the top leg positionedon a smooth, low-friction surface to allowdetection of slight movements with minimal in-fluence of gravity (Supplemental Figure 1E).Tasks included attempts to continuouslymodulate leg muscle activity while followinga digitally displayed sinusoidal wave, as wellas volitionally increasing muscle activity whencued by a 3-step audio tone. Visual feedbackwas provided by mirrors reflecting the lowerlimbs and by real-time streaming surface elec-tromyography (EMG).

Data ProcessingMuscle activity was recorded via surface EMGelectrodes (LabChart and PowerLab, ADIn-struments). Data were sampled at 4 kHz andwere exported from LabChart software andanalyzed using MATLAB software (The Math-Works Inc). The EMG data were filtered usinga 60-Hz notch filter and a bandpass filter of 20

016/j.mayocp.2017.02.014 545

http://www.neurorti.com

http://www.neurorti.com




22 wk

Motor training(61 sessions)

EES motor training(8 sessions)

2 wk

L RF

Lumbosacral spinally evokedmotor potentials in R SOL

Conditioning stimulation

L VLL MHL TAL MGL SOLR RFR VLR MHR TAR MGR SOLR FCR

100 µV

Enrollment

Presurgical 8 wk motor trainingE

FDC

8 wk

AUC

(% m

axim

um)

16 wk

16 wk motor training

Postsurgical

10 ms200 µV

100 µV

100 µV

25 µV

50 ms

L TA

L RF

L VLL M

HL T

AL M

GL S

OL

L RF

L VLL M

HL T

AL M

GL S

OL

L MG

L SOL

5 s100

50

0

Motor

Pinprick

Light touch

Enrollment

Enrollment

Presurgical

Presurgical

Postsurgical

Postsurgical

A

B

Enrollmentassessment

Surgicalimplantation Surgical

recovery

3 wk

Res

pons

e la

tenc

y (m

s) 25

20

NR

print&

web4C=FPO

FIGURE 1. The patient’s spinal cord injury (SCI) profile. A, Experimental timeline. Assessments were performed at enrollment, beforesurgery, and after 3 weeks of surgical recovery. B, Shaded regions depict American Spinal Injury Association motor, pinprick, and lighttouch scores across assessments. C, Evoked potential latencies were within reference ranges when recorded over the spine below thelevel of injury during bilateral tibial nerve stimulation. Green squares and red diamonds represent right and left tibial nerve stimulation,respectively. There were no detectable responses at the cortical level (black x’s and blue circles). D, Motor evoked potentials were notobserved from any leg muscle during transcranial magnetic stimulation of the motor cortex. Recordings from the right arm flexor carpiradialis (R FCR) indicate that stimulation intensity adequately evoked responses. Bold trace represents an average of 5 responses. E,Conditioning stimuli applied to the motor cortex, ulnar nerve, or cervical spine were followed by a stimulus delivered to thelumbosacral spinal cord. Conditioned depression or potentiation of spinally evoked motor potentials was not observed. Tracesrepresent an average of 5 responses. Blue vertical lines indicate timing of conditioning stimuli before the onset of lumbosacral stimuli(black dashed line). Top and bottom traces are unconditioned evoked potentials. F, Evidence of discomplete SCI and progressiveincreases in muscle activation observed at weeks 8 and 16 of presurgical motor training during 2 attempts (white and gray panels) ofmaximal volitional contraction. Red dashed lines indicate start of volitional attempt. Mean � SD area under the curve (AUC) wasextracted from the root mean square envelope of rectified electromyographic recordings from left leg muscles and averaged for the 2trials. L ¼ left limb; MG ¼ medial gastrocnemius; MH ¼ medial hamstrings; ms ¼ milliseconds; NR ¼ no response; R ¼ right limb;RF ¼ rectus femoris; SOL ¼ soleus; TA ¼ tibialis anterior; mV ¼ microvolts; VL ¼ vastus lateralis.

MAYO CLINIC PROCEEDINGS

546

to 1000 Hz. Additional analyses included full-wave rectification on the filtered EMG signalafter subtraction of the mean backgroundsignal, and then the root mean square of the

Mayo Clin Proc. n April 2017

EMG activity was calculated as the squareroot of the mean square averaged over a win-dow length of 4000 samples (sliding overlapof 3000 samples).

;92(4):544-554 n http://dx.doi.org/10.1016/j.mayocp.2017.02.014www.mayoclinicproceedings.org




EES System ImplantationAfter the 61 presurgical motor trainingsessions, an EES device (RestoreSensor Sure-Scan MRI, Medtronic) was surgicallyimplanted and connected to a 16-contactelectrode array (Specify 5-6-5, Medtronic)positioned on the dorsal epidural surface ofthe lumbosacral spinal cord. Intraoperativeradiography (Supplemental Figure 2, availableonline at http://www.mayoclinicproceedings.org) and epidurally evoked motor responses(0-5.5 V, 0.5 milliseconds, 1 Hz) were usedto confirm array position over the lumbosacralenlargement of the spinal cord.23,24

Identification of EES Parameters ThatEnable Volitional Control of Motor FunctionsAfter 3 weeks of postsurgical recovery, voli-tional activation of leg muscles was attemptedin the presence of EES while the patient waspositioned side-lying or upright (with andwithout body weight support). The EES pa-rameters that enabled voluntary control ofrhythmic lower limb muscle activity wereidentified while in the side-lying positionwith the top leg suspended using nonelasticnylon fabric support slings to allow free limbmovement.25,26 While the patient was in anupright position, with trainer assistance orwhile in a body weight support harness, EESsettings were identified that allowed either in-dependent standing or volitional control ofsteplike muscle activity and limb movements.

Over 2 weeks (8 sessions, 5-7 hours each),active electrode configurations and stimulationparameters were adjusted to allow volitionalcontrol of the muscles of interest. A systematicapproach was used to determine the bestparameter settings. Previous reports werereviewed to provide starting points for stimu-lation.8,10,11,27,28 We chose frequencies (25and 40 Hz for volitional control and steppingand 15 Hz for standing) and a pulse duration(0.21 milliseconds) based on previous reports.The electrode configuration was adjustedbased on an algorithm where we initially eval-uated the effect of wide-field (via electrodesmost distal from each other) vs local-field(via electrodes most proximal to each other)stimulation, with both cathode and anodeelectrodes tested (ie, reversing polarity) foreach configuration. The effect of right vs left


lateralized electrode location was also assessed.For each configuration, we incrementallyincreased voltage intensity from 0 to 6 V.Voltage intensity was reduced to a comfortablelevel or was turned off if the patient reportedany signs of discomfort. Once volitional con-trol was achieved, that EES setting was heldconstant for further repetitions of the intendedtask.

RESULTS

Clinical Assessment OutcomesPresurgical and postsurgical clinical assess-ments showed no change in motor or sensoryscores on the AIS (Figure 1B). Tibial somato-sensory evoked potentials showed normal la-tencies at peripheral and lumbar spine sites,but there were no detectable cortical responses(Figure 1C). Transcranial magnetic stimulationelicited normal motor evoked potentials in theupper extremities, but there were no responsesobserved from lower extremity muscles(Figure 1D). These assessments were per-formed on several occasions during the preeand posteEES implant periods with nochanges identified.

Electrophysiologic Assessment ofTranslesional Neural ConnectivityThere were no significant differences betweenspinally evoked motor potentials that wereelicited after conditioning and those evoked in-dependent of conditioning pulses (Figure 1E).In summary, neural connectivity across theinjury site was not detectable during clinicalor electrophysiologic examinations at the timeof enrollment into the study, before EES surgi-cal implantation, or after surgery.

Motor Training OutcomesThroughout the presurgical motor trainingsessions, no volitional EMG modulation orcontrol of lower limb movement wasobserved. When in side-lying, some prolongedattempts (lasting 10-15 seconds) of maximalvolitional flexion of all joints of the leg whilecontracting the anterior trunk musclesresulted in nonspecific coactivation of agonistand antagonist lower leg muscles over thecourse of motor task training (Figure 1F).This EMG activity was first recognizable inthe second month of motor training and was

016/j.mayocp.2017.02.014 547






548

observed on multiple testing sessions over thecourse of motor training before surgery. TheEMG recordings showed delayed muscle acti-vation that occurred in a generalized, co-contractile mass activation pattern. The patientdid not volitionally target the muscles acti-vated below the level of injury; rather, thisobserved activity was the result of his maximalefforts to flex all the joints of his leg and con-tract anterior trunk muscles. This observationis similar to descriptions in the literature of adiscomplete SCI profile.2-5,29

Intraoperative Positioning of the EpiduralElectrode ArrayIntramuscular EMG recordings were gatheredfrom select lower limbmuscles. Spinally evokedmotor potentials via the implanted epiduralelectrode were used to position the array toallow recruitment of either distal or proximallower limb musculature bilaterally. Recruit-ment of distal or proximal muscles at lowvoltage intensity depended on active electrodelocation (Supplemental Figure 3, availableonline at http://www.mayoclinicproceedings.org). At high voltage intensity, coactivation ofboth distal and proximal leg muscles wasobserved.

EES Enabled Volitional Control ofTask-Specific Muscle ActivityIn a side-lying position with the top leg posi-tioned on a smooth, low-friction surface, stim-ulation intensities that were at the thresholdfor evoking nonvolitional muscle activitywere identified for each EES configuration.The EES at subthreshold intensities, in whichstimulation alone did not directly elicit muscleactivity, facilitated volitional production ofmotor activity only when the patient attemp-ted to move his leg (Figure 2A, left panel).At EES intensities that were at the thresholdfor eliciting muscle activity, the EMG ampli-tude of muscles involved in the task per-formed increased, but limb movements wereminimal (eg, flexor muscles such as the medialhamstrings could be activated during kneeflexion attempts, but minimal knee flexionwas observed) (Figure 2A, center panel). TheEES settings that facilitated volitional controlof movement enabled control of musclegroups involved in producing joint-specificmovement (Figure 2A, right panel)


(Supplemental Video 1, available online athttp://www.mayoclinicproceedings.org). Thisvolitional control could be modulated in anincreasing and decreasing manner using eithervisual cuing that consisted of a digitally dis-played scrolling sinusoidal wave (Figure 2B)or audio cuing as a tone consisting of 3increasing intensities (Figure 2C). In contrast,without EES, muscle activity was either absentor unorganized and could not be controlledwith respect to timing, intensity, or duration(Figure 2B and C).

EES Enabled Volitional Control of RhythmicMotor Activity to Produce SteplikeMovements in the Side-Lying PositionThe EES settings that enabled volitional rhyth-mic activity were also identified (Figure 3A).Lateralizing electrode configurations facilitatedipsilateral control of rhythmic activity, andshifting active electrodes contralaterally onthe array attenuated EES facilitation of ipsilat-eral EMG activity (Figure 3B). Once a subset ofparameters was identified that enabled voli-tional control of rhythmic activity, with thetop limb suspended using a fabric supportsling, the patient was able to initiate andthen modulate the amplitude of steplikemovements, as well as volitionally terminaterhythmic steplike activity (SupplementalVideo 2 and Supplemental Figure 4, availableonline at http://www.mayoclinicproceedings.org).

EES Enabled Independent StandingWithout EES, the patient was unable to standwithout trainer assistance. In the presence ofEES, he was able to stand independently formore than 1.5 minutes, using his arms tomaintain balance (Figure 4) (SupplementalVideo 3, available online at http://www.mayoclinicproceedings.org). While indepen-dently standing, he was able to shift his weightforward, backward, and laterally using hisarms to assist. During standing sessions per-formed in the presence of EES, multiple fac-tors contributed to the patient requesting abrief break from standing, including changesin blood pressure and heart rate when transi-tioning from sitting to standing, arm musclefatigue from maintaining balance, and cardio-vascular fatigue. The duration of standingwas not limited by a decrease in efficacy of











No EES L TA

3 trial AVG

EES L TA

EES L MG

EES L SOL

3 trial AVG

No EES L MG

No EES L SOL

3 trial AVG

3 trial AVG

3 trial AVG

3 trial AVG

No EES L TA

3 trial AVG

EES L TA

EES L MG

EES L SOL

3 trial AVG

No EES L MG

No EES L SOL

3 trial AVG

3 trial AVG

3 trial AVG

3 trial AVG

10 µV

5 µV

5 µV

L MH

L TA

SubthresholdEES

L RF

L VL

L M

HL

TAL

MG

L SO

L

100

50

0

Motor thresholdEES

Volitional legflexion enabled

L MH

L TA

L MH

L TA

Visual cue Audio cue

5 µV

5 µV

10 µV

10 µV

1 µV

10 µV

40 Hz210 µs5.2 V

25 Hz210 µs

2 V

40 Hz210 µs3.8 V

40 Hz210 µs5.2 V

40 Hz210 µs5.2 V

1 µV

0.5 µV

5 µV

4 s

1 s

25 µV

50 µV

1 s

50 µV

50 µV

1 s

5 s

1 µV

2 µV

10 µV

5 µV

10 µV

10 µV

20 µV

5 µV

5 µV

1 µV

0.5 µV

10 µV

1 µV

5 µV

AU

C (%

max

imum

)

L RF

L VL

L M

HL

TAL

MG

L SO

L

100

50

0

AU

C (%

max

imum

)

L RF

L VL

L M

HL

TAL

MG

L SO

L

100

50

0

AU

C (%

max

imum

)

A

B C

print&web4C=FPO

FIGURE 2. Volitional control of task-specific motor activities with and without epidural electrical stimulation (EES). A, Electromyo-graphic (EMG) recordings are shown during attempts to volitionally perform leg flexion in the presence of EES using voltage intensitiesthat were subthreshold of motor activity (left panel), at threshold of motor activity (center panel), and that enabled volitionalmovement (right panel) during the first 2 weeks of EES. Blue shaded boxes indicate attempts to volitionally perform left knee flexion.Area under the curve (AUC) values of rectified EMG of left leg muscles normalized to the maximal response recorded from themedial hamstring in the presence of EES that enabled volitional leg flexion. The EMG recordings during attempts to volitionallycontract lower limb muscles in the presence of EES with a visual display of a sinusoidal cue (B) or an audio cue (C) consisting of 3increasing intensities without EES (red plots) and with EES (black plots). Rectified EMG for the first trial (upper panels) and root meansquare averages (AVGs) of 3 trials are shown with standard deviations (shaded regions). For all the panels, EES electrode configu-rations during the task performed are displayed as cathodes (black boxes) and anodes (red boxes), with stimulation settings shownbelow each array. Hz ¼ hertz; L ¼ left; MG ¼ medial gastrocnemius; MH ¼ medial hamstrings; ms ¼ milliseconds; RF ¼ rectusfemoris; SOL ¼ soleus; s ¼ seconds; TA ¼ tibialis anterior; ms ¼ microseconds; mV ¼ microvolts; VL ¼ vastus lateralis.


stimulation during a session. However, asstimulation parameters were adjusted to findsettings that enabled independent standing,the patient did report bouts of trunk pares-thesia, trunk muscle discomfort, and feelingshort of breath, which required brief periodsof rest lasting approximately 5 minutes withEES turned off.

EES Enabled Steplike Movements in theUpright PositionWhile in an upright position, with bodyweight partially supported, the participantwas able to voluntarily generate leg flexion of1 leg with coordinated contralateral leg


extension in the presence of continuous EES.He was able to volitionally alternate this activ-ity in the legs to produce steplike movements(Figure 5, Supplemental Video 4, availableonline at http://www.mayoclinicproceedings.org).

DISCUSSIONIn this report, a patient diagnosed as havingsensory and motor complete (AIS-A) SCI wasidentified to have a discomplete SCI duringmotor task training before EES implantation.After EES implantation and surgical recovery,EES enabled (1) volitional control of task-specific lower limb muscle activity in a

016/j.mayocp.2017.02.014 549





100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

100 µV

L RF

L VL

L MH

L TA

L MG

L SOL

L RF

L VL

L MH

L TA

L MG

L SOL

2 V 3 V 4 V 5 V 6 V

25 Hz

210 µs

25 Hz210 µs5.5 V

25 Hz210 µs5.5 V

1 s 1 s 1 s

A B

print&web4C=FPO

FIGURE 3. Identification of epidural electrical stimulation (EES) parameters that enable rhythmic motor activity. A, Left leg elec-tromyography (EMG) while increasing EES from 2 to 6 V is shown with the patient lying on his right side with his left leg elevated.From 3 to 5 V the patient was unable to initiate rhythmic muscle activity. At 6 V, EES enabled control of rhythmic activity. B, Activeelectrodes positioned laterally enabled rhythmic activity in the limb ipsilateral to stimulation (left plot). Shifting active electrodes to thecontralateral side of the array did not facilitate rhythmic EMG in the ipsilateral leg. The EES electrode configurations during the taskperformed are displayed as cathodes (black boxes) and anodes (red boxes), with stimulation settings shown below. Hz ¼ hertz;L ¼ left; MG ¼ medial gastrocnemius; MH ¼ medial hamstrings; RF ¼ rectus femoris; SOL ¼ soleus; TA ¼ tibialis anterior; ms ¼microseconds; mV ¼ microvolts; V ¼ volts; VL ¼ vastus lateralis.


550

side-lying position; (2) voluntary initiation,modulation, and termination of rhythmicmuscle activity to produce steplike move-ments while side-lying; (3) full weightbearingindependent standing; and (4) volitional con-trol of muscle activity to produce steplikemovements while in a vertical position withbody weight partially supported. For the firsttime, all of these functions, which were absentbefore EES, were enabled in the presence ofEES during the first 2 weeks (8 sessions total)in the same participant. During these trainingsessions, in the presence of EES, the patientregained volitional control of paralyzed mus-cles and was able to control rhythmic steplikemotor activity, both in side-lying and whilestanding with body weight partially sup-ported. In addition, EES facilitated indepen-dent standing using his arms placed onsupport bars for balance only. Without EES,the patient was unable to perform any of thesetasks.

Early reports of evidence of central patterngenerators in the human spinal cord led to thedevelopment of strategies to engage spinal net-works that drive locomotor activity afterSCI.6,30-34 Since then, numerous studies haveshown that step training on a treadmill can


improve locomotor performance after incom-plete SCI.35-38 Several reports of EES of thelumbosacral enlargement after SCI previouslydescribed that locomotor spinal networkscould be facilitated to produce nonvolitionalsteplike patterns of limb movement withrhythmic muscle activity.6-8 A later studydescribed a series of 4 cases of chronic lossof motor function due to SCI, where EES com-bined with motor rehabilitation enabled voli-tional control of motor activity in the lowerlimbs.11 However, EES-enabled independentstanding did not occur for at least 17 weeks.In addition, to date, there are no reports inthe literature of recovery of volitional controlof EES-facilitated rhythmic activity to allowinitiation, modulation, and termination ofsteplike activity and volitional control oftask-specific muscle activity (eg, flexion andextension) in the same patient.

The results of this case report, combinedwith previous reports of spinal cord neuromo-dulation restoring function after motor com-plete paralysis,10,11,25,39 suggest thatsubfunctional neural connections are likelypresent in some cases of clinically motor com-plete (AIS-A or AIS-B) SCI, and these neuralconnections can be identified and used for




100

L RF

L VL

L MH

L TA

L MG

L SOL

R RF

R VL

R MH

R TA

R MG

R SOLL

RFL

VLL

MH

L TA

L M

GL

SOL

R RF

R VL

R M

HR

TAR

MG

R SO

L

L RF

L VL

L M

HL

TAL

MG

L SO

LR

RFR

VLR

MH

R TA

R M

GR

SOL

L RF

L VL

L M

HL

TAL

MG

L SO

LR

RFR

VLR

MH

R TA

R M

GR

SOL

50

0

AU

C (%

max

imum

)

15 Hz210 µs2.2 V

15 Hz210 µs1.5 V

300 µV

5 s

EES off Motor threshold EES Independent standing enabled

print&

web4C=FPO

FIGURE 4. Epidural electrical stimulation (EES)efacilitated independent standing. Without EES, trainer support was required at thepelvis and knees, with the patient using arm support to maintain upright posture. Electromyographic (EMG) recordings from lowerlimb muscles show no muscle activity during trainer-assisted standing without EES. During threshold EES (1.5 V, 15 Hz, 210 mi-croseconds), trainer assistance was required at the knees with minimal support at the pelvis. The EMG records indicate activity in theleft vastus lateralis (L VL) and bilaterally in the medial hamstrings. At increased EES voltage intensity (2.2 V, 15 Hz, 210 microseconds),independent standing was achieved for more than 1.5 minutes. Trainers provided no assistance but remained in position to assist ifneeded. Active electrode positions on the EES array are displayed as cathodes (black) and anodes (red). Bottom panels show areaunder the curve (AUC) values from rectified EMG as percentage maximum of the left vastus lateralis during independent standing.Black bars represent EMG activity in microvolts and time in seconds. Hz ¼ hertz; L ¼ left; MG ¼ medial gastrocnemius; MH ¼ medialhamstrings; R ¼ right; RF ¼ rectus femoris; s ¼ seconds; SOL ¼ soleus; TA ¼ tibialis anterior; mV ¼ microvolts; V ¼ volts.


enabling volitional control of motor functionvia EES. In this study, we used multiple tech-niques to attempt to diagnose residual fibersacross the injury level; however, no detectable


connections were observed. Still, before EES,nonspecific muscle activity (ie, coactivationof agonist and antagonist lower leg muscles)during maximal volitional flexion of all joints

016/j.mayocp.2017.02.014 551



L EXTL FLX

L RF

L VL

L MH

L TA

L MG

L SOL

R EXTR FLX

R RF

R VL

R MH

R TA

R MG

R SOL

25 Hz210 µs5.5 V

100 µV1 s

100 µV1 s

print&web4C=FPO

FIGURE 5. Bilateral steplike electromyographic (EMG) activity in an upright position with partial bodyweight supported. The EMG recordings from the left and right legs are shown for 5 cycles of volitionalextension and flexion bilaterally in the presence of epidural electrical stimulation (EES). Active electrodepositions on the EES array are displayed as cathodes (black) and anodes (red). Red and blue barsrepresent the period when the patient was attempting leg extension and flexion activity. EXT ¼ legextension; FLX ¼ leg flexion; Hz ¼ hertz; L ¼ left; MG ¼ medial gastrocnemius; MH ¼ medial hamstrings;r ¼ right; RF ¼ rectus femoris; SOL ¼ soleus; TA ¼ tibialis anterior; ms ¼ microseconds; mV ¼ microvolts;V ¼ volts; VL ¼ vastus lateralis.


552

of the leg and simultaneous contraction ofanterior trunk muscles was observed, indi-cating that this patient has discomplete SCI.At the same time, the observed EMG activityincreased in amplitude during presurgical mo-tor training, suggesting that motor trainingmay have influenced the overall excitabilityof spinal motor circuitry via unidentified resid-ual fibers that cross the injury site.

Although the present data suggest that pa-tients found to have discomplete SCI could bepotential candidates for EES therapy, werecognize that the indication of discompleteSCI currently provides only general informa-tion about the injury status and lacks a specificdefinition that characterizes specific ascendingor descending spinal pathways. This lack of adistinct definition is due to limited diagnosticapproaches that can be used to identify thepathways transmitting residual, yet subfunc-tional, descending and ascending signalsacross the injury in AIS-A patients. Therefore,


the neural substrates underlying the phenom-enon of discomplete SCI and the extent towhich these substrates contribute to EES-enabled functional recovery described in thepresent study remains to be determined.

In summary, these results provide evidenceof successful replication of a previous clinicaltrial using EES therapy and altogether providefurther evidence that spinal cord neuromodula-tion strategies combined with intense motorrehabilitation can facilitate functional recoveryeven when initiated years after the occurrenceof a motor complete spinal injury.

CONCLUSIONThis report provides evidence of successfulreplication of previously reported results that(1) EES enabled volitional control of motor ac-tivity and (2) EES enabled independent stand-ing after several months of training in 2 AIS-Aand 2 AIS-B patients. In addition, this is thefirst report of EES-facilitated volitional control





of task-specific motor activity and control ofrhythmic steplike movements in a single pa-tient with clinically complete chronic SCI dur-ing the first 2 weeks of EES.

ACKNOWLEDGMENTWe thank Penelope S. Duffy, PhD, for editorialexpertise; Beth A. Cloud, PhD, DPT, Daniel D.Veith, MS, Deborah E. Hare, MS, and YulunLi, BS, for assistance during motor task training;Andrew Schmeling, BSEE, for guidance on stim-ulation programming; Carlos Cuellar Ramos,PhD, and Riazul Islam, MS, for assistance withdata processing; and Amanda Turner, BS,Cynthia J. Stoppel, AAS, Elizabeth M. Mosier,BA, Marc N. Shaft, BA, and Terri A. Gardnerfor administrative support.

Drs Grahn and Lavrov contributed equallyto this work.

SUPPLEMENTAL ONLINE MATERIALSupplemental material can be found online athttp://www.mayoclinicproceedings.org. Sup-plemental material attached to journal articleshas not been edited, and the authors take re-sponsibility for the accuracy of all data.

Abbreviations and Acronyms: AIS = American SpinalInjury Association Impairment Scale; AUC = area under thecurve; AVG = root mean square average; EES = epiduralelectrical stimulation; EMG = electromyography; EXT = legextension; FLX = leg flexion; L = left limb; MG = medialgastrocnemius; MH = medial hamstrings; NR = no response;R = right limb; RF = rectus femoris; R FCR = right flexorcarpi radialis; SCI = spinal cord injury; SOL = soleus; TA =tibialis anterior; VL = vastus lateralis

Grant Support: This study was supported by the BroccoliFoundation, the Christopher and Dana Reeve Foundation,Mayo Clinic Center for Clinical and Translational Sciences,Mayo Clinic Rehabilitation Medicine Research Center,Mayo Clinic Transform The Practice, the Craig H. NeilsenFoundation, the Bel13ve in Miracles Foundation, and TheGrainger Foundation.

Potential Competing Interests: Drs Gad, Gerasimenko,and Edgerton are shareholders in NeuroRecovery Technol-ogies; Dr Edgerton is president and chair of the company’sboard of directors; and Drs Gad, Gerasimenko, and Edge-rton hold investorship rights on intellectual propertylicensed by the regents of the University of California toNeuroRecovery Technologies and its subsidiaries.

Correspondence: Address to Kristin D. Zhao, PhD, orKendall H. Lee, MD, PhD, Mayo Clinic, 200 First St SW,Rochester, MN 55905 ([email protected]; [email protected]).


REFERENCES1. Kakulas BA. Pathology of spinal injuries. Cent Nerv Syst Trauma.

1984;1(2):117-129.2. Sherwood AM, Dimitrijevic MR, McKay WB. Evidence of sub-

clinical brain influence in clinically complete spinal cord injury:discomplete SCI. J Neurol Sci. 1992;110(1-2):90-98.

3. Tuel S, McKay WB, Dimitrijevic MR. The discomplete syn-drome: clinical implications of residual suprasegmental influenceacross a clinically complete spinal cord injury. Arch Phys MedRehabil. 1992;73:1.

4. Dimitrijevic MR, Hsu CY, McKay WB. Neurophysiologicalassessment of spinal cord and head injury. J Neurotrauma.1992;9(suppl 1):S293-S300.

5. McKay WB, Lim HK, Priebe MM, Stokic DS, Sherwood AM.Clinical neurophysiological assessment of residual motor con-trol in post-spinal cord injury paralysis. Neurorehabil NeuralRepair. 2004;18(3):144-153.

6. Dimitrijevic MR, Gerasimenko Y, Pinter MM. Evidence for a spi-nal central pattern generator in humans. Ann N Y Acad Sci.1998;860:360-376.

7. Minassian K, Jilge B, Rattay F, et al. Stepping-like movements inhumans with complete spinal cord injury induced by epiduralstimulation of the lumbar cord: electromyographic study ofcompound muscle action potentials. Spinal Cord. 2004;42(7):401-416.

8. Danner SM, Hofstoetter US, Freundl B, et al. Human spinal lo-comotor control is based on flexibly organized burst genera-tors. Brain. 2015;138(pt 3):577-588.

9. Jilge B, Minassian K, Rattay F, et al. Initiating extension of thelower limbs in subjects with complete spinal cord injury byepidural lumbar cord stimulation. Exp Brain Res. 2004;154(3):308-326.

10. Harkema S, Gerasimenko Y, Hodes J, et al. Effect of epiduralstimulation of the lumbosacral spinal cord on voluntary move-ment, standing, and assisted stepping after motor completeparaplegia: a case study. Lancet. 2011;377(9781):1938-1947.

11. Angeli CA, Edgerton VR, Gerasimenko YP, Harkema SJ.Altering spinal cord excitability enables voluntary movementsafter chronic complete paralysis in humans. Brain. 2014;137(pt 5):1394-1409.

12. Andrews JC, Stein RB, Roy FD. Reduced postactivation depres-sion of soleus H reflex and root evoked potential after transcra-nial magnetic stimulation. J Neurophysiol. 2015;114(1):485-492.

13. Andrews JC, Stein RB, Roy FD. Post-activation depression inthe human soleus muscle using peripheral nerve and transcuta-neous spinal stimulation. Neurosci Lett. 2015;589:144-149.

14. Roy FD, Bosgra D, Stein RB. Interaction of transcutaneous spi-nal stimulation and transcranial magnetic stimulation in humanleg muscles. Exp Brain Res. 2014;232(6):1717-1728.

15. Sayenko DG, Atkinson DA, Floyd TC, et al. Effects of pairedtranscutaneous electrical stimulation delivered at single anddual sites over lumbosacral spinal cord. Neurosci Lett. 2015;609:229-234.

16. Barthélemy D, Willerslev-Olsen M, Lundell H, Biering-Sorensen F, Nielsen JB. Assessment of transmission in specificdescending pathways in relation to gait and balance followingspinal cord injury. Prog Brain Res. 2015;218:79-101.

17. Squair JW, Bjerkefors A, Inglis JT, Lam T, Carpenter MG.Cortical and vestibular stimulation reveal preserved descendingmotor pathways in individuals with motor-complete spinal cordinjury. J Rehabil Med. 2016;48(7):589-596.

18. Ilic TV, Pötter-Nerger M, Holler I, et al. Startle stimuli exertopposite effects on human cortical and spinal motor systemexcitability in leg muscles. Physiol Res. 2011;60(suppl 1):S101-S106.

19. Zehr EP, Frigon A, Hoogenboom N, Collins DF. Facilitation ofsoleus H-reflex amplitude evoked by cutaneous nerve stimula-tion at the wrist is not suppressed by rhythmic arm movement.Exp Brain Res. 2004;159(3):382-388.

016/j.mayocp.2017.02.014 553


mailto:[email protected]






554

20. Frigon A, Collins DF, Zehr EP. Effect of rhythmic arm movementon reflexes in the legs: modulation of soleus H-reflexes and so-matosensory conditioning. J Neurophysiol. 2004;91(4):1516-1523.

21. Wolfe DL, Hayes KC, Potter PJ, Delaney GA. Conditioninglower limb H-reflexes by transcranial magnetic stimulation ofmotor cortex reveals preserved innervation in SCI patients.J Neurotrauma. 1996;13(6):281-291.

22. Harkema SJ, Schmidt-Read M, Behrman AL, Bratta A, Sisto SA,Edgerton VR. Establishing the NeuroRecovery Network: multi-site rehabilitation centers that provide activity-based therapiesand assessments for neurologic disorders. Arch Phys Med Reha-bil. 2012;93(9):1498-1507.

23. Kendall FP, McCreary EK, Provance PG, Rodgers M. Muscles:Testing and Function, With Posture and Pain. Baltimore, MD: Wil-liams & Wilkins; 1993.

24. Sayenko DG, Angeli C, Harkema SJ, Edgerton VR,Gerasimenko YP. Neuromodulation of evoked muscle poten-tials induced by epidural spinal-cord stimulation in paralyzed in-dividuals. J Neurophysiol. 2014;111(5):1088-1099.

25. Gerasimenko Y, Lu D, Modaber M, et al. Noninvasive reactiva-tion of motor descending control after paralysis. J Neurotrauma.2015;32(24):1968-1980.

26. Gerasimenko Y, Gorodnichev R, Moshonkina T, Sayenko D,Gad P, Reggie Edgerton V. Transcutaneous electrical spinal-cord stimulation in humans. Ann Phys Rehabil Med. 2015;58(4):225-231.

27. Rejc E, Angeli C, Harkema S. Effects of lumbosacral spinal cordepidural stimulation for standing after chronic complete paraly-sis in humans. PLoS One. 2015;10(7):e0133998.

28. Rejc E, Angeli CA, Bryant N, Harkema S. Effects of stand andstep training with epidural stimulation on motor function forstanding in chronic complete paraplegics [published onlineOctober 5, 2016]. J Neurotrauma. http://dx.doi.org/10.1089/neu.2016.4516.

29. Sherwood AM, McKay WB, Dimitrijevic MR. Motor control af-ter spinal cord injury: assessment using surface EMG. MuscleNerve. 1996;19(8):966-979.


30. Roby-Brami A, Bussel B. Long-latency spinal reflex in man af-ter flexor reflex afferent stimulation. Brain. 1987;110(pt 3):707-725.

31. Bussel B, Roby-Brami A, Néris OR, Yakovleff A. Evidencefor a spinal stepping generator in man. Paraplegia. 1996;34(2):91-92.

32. Bussel B, Roby-Brami A, Azouvi P, Biraben A, Yakovleff A,Held JP. Myoclonus in a patient with spinal cord transection:possible involvement of the spinal stepping generator. Brain.1988;111(pt 5):1235-1245.

33. Fung J, Stewart JE, Barbeau H. The combined effects of cloni-dine and cyproheptadine with interactive training on the mod-ulation of locomotion in spinal cord injured subjects. J NeurolSci. 1990;100(1-2):85-93.

34. Calancie B, Needham-Shropshire B, Jacobs P, Willer K, Zych G,Green BA. Involuntary stepping after chronic spinal cord injury:evidence for a central rhythm generator for locomotion in man.Brain. 1994;117(pt 5):1143-1159.

35. Wernig A, Müller S. Laufband locomotion with body weightsupport improved walking in persons with severe spinal cordinjuries. Paraplegia. 1992;30(4):229-238.

36. Harkema SJ, Hurley SL, Patel UK, Requejo PS,Dobkin BH, Edgerton VR. Human lumbosacral spinalcord interprets loading during stepping. J Neurophysiol.1997;77(2):797-811.

37. Alexeeva N, Sames C, Jacobs PL, et al. Comparison of trainingmethods to improve walking in persons with chronic spinalcord injury: a randomized clinical trial. J Spinal Cord Med.2011;34(4):362-379.

38. Harkema SJ, Hillyer J, Schmidt-Read M, Ardolino E, Sisto SA,Behrman AL. Locomotor training: as a treatment of spinalcord injury and in the progression of neurologic rehabilitation.Arch Phys Med Rehabil. 2012;93(9):1588-1597.

39. Lu DC, Edgerton VR, Modaber M, et al. Engaging cervicalspinal cord networks to reenable volitional control ofhand function in tetraplegic patients. Neurorehabil NeuralRepair. 2016;30(10):951-962.


http://dx.doi.org/10.1089/neu.2016.4516

http://dx.doi.org/10.1089/neu.2016.4516



enabling task-specific volitional motor functions via spinal cord … · 2017. 8. 8. · functions...

Documents