← Back

SCS for Post-Stroke Hemiparesis

Spinal Cord Stimulation Improves Motor Function and Spasticity in Chronic Post-Stroke Upper Limb Hemiparesis

Share Share Copied! Compare

Year of Publication: 2025

Authors: Roberto M. de Freitas, Shovan Bhatia, Erynn Sorensen, ..., Marco Capogrosso

Journal: medRxiv

Citation: medRxiv preprint doi: https://doi.org/10.1101/2025.08.01.25332445

Link: https://doi.org/10.1101/2025.08.01.25332445

PDF: https://pmc.ncbi.nlm.nih.gov/articles/PM....25332445v1.pdf

Clinical Question

Can cervical epidural spinal cord stimulation safely and effectively improve motor function and reduce spasticity in individuals with chronic post-stroke upper-limb hemiparesis?

Study Overview

Objective

To assess safety and preliminary efficacy of cervical epidural spinal cord stimulation (SCS) for chronic post-stroke upper-limb hemiparesis

Study Summary

SCS immediately improved motor function (+32% strength, +5.6 FMA points)
3/7 participants with residual corticospinal connectivity regained hand/finger function
All participants improved by +6.6 FMA points at study end with decreased spasticity

Intervention

Cervical epidural spinal cord stimulation using two 8-contact leads (40-100Hz, 0.2-8mA, 200-400μs pulse width)

Patients per Arm

7 total participants (all received both conditions)

Bottom Line

Cervical SCS is safe and provides immediate assistive improvements in strength, dexterity, and spasticity, with therapeutic effects emerging over 4 weeks, regardless of impairment severity. Spared sensory function may predict responsiveness.

        Major Points
        First-in-human pilot trial (NCT04512690) testing cervical epidural SCS in 7 participants with chronic hemiparesis (FMA 15-35)
Two percutaneous 8-contact leads implanted unilaterally in cervical epidural space (C3-T1) for 4 weeks
Immediate assistive effects: average +32% strength increase and +5.6 FMA points with SCS ON
3/7 participants with residual corticospinal connectivity (MEP+) regained hand/finger function
Therapeutic effects emerged despite only 8.6 hours of motor activity (5.5 hours with SCS ON)
Participants improved by average +6.6 FMA points at study end compared to baseline
Spasticity decreased in all participants (MAS scores)
No serious adverse events occurred; 14 mild AEs documented
Spared sensory function may be a determinant of SCS responsiveness
85% of maximum FMA improvement achieved by week 2

      

Design

Study Type: Non-randomized, open-label, prospective pilot study

Randomization:

Blinding: Evaluator blinded for FMA assessments; participants served as their own controls (SCS ON vs OFF)

Enrollment Period: Not specified

Follow-up Duration: 4 weeks (with implant) + 1 month follow-up post-explant

Centers: 1

Countries: USA

Sample Size: 7

Analysis: Bootstrap analysis with 10,000 samples for paired comparisons; 95% confidence intervals; Pearson correlations for EMG-kinematic relationships

Inclusion Criteria

Ages 21-70 years
Ischemic or hemorrhagic stroke >6 months prior
Upper limb hemiparesis with baseline FMA scores 7-40
Passed rigorous medical evaluation
No severe comorbidities
No previously implanted medical devices
Not claustrophobic
Not pregnant or breastfeeding
Not receiving anticoagulant, anti-spasticity, or anti-epileptic medications during study

Exclusion Criteria

Severe comorbidities
Previously implanted medical devices
Claustrophobia
Pregnancy or breastfeeding
Receiving anticoagulant medications
Receiving anti-spasticity medications
Receiving anti-epileptic medications

Baseline Characteristics

Characteristic	Control	Active
Note	Within-subject design - all participants served as their own controls
N	7
Age range	30-70 years
Sex	4 female, 3 male
Years post-stroke	2-10 years (range)
Stroke type	3 hemorrhagic, 4 ischemic
Baseline FMA motor	15-35 (range), mean ~26.3
Baseline FMA sensory	2-12 (range)
MEP status	3 MEP+, 3 MEP-, 1 likely MEP+
SSEP status	5 SSEP+, 2 SSEP-
Hand opening ability	3 yes, 4 no
Same participants - within-subject crossover design		SCS ON condition

Arms

Field	Control	SCS ON
Intervention	No spinal cord stimulation - baseline motor function assessment	Cervical epidural spinal cord stimulation via two 8-contact percutaneous leads (C3-T1 levels) positioned lateral to spinal midline, targeting ipsilateral dorsal roots. Parameters: 40-100Hz frequency, 0.2-8mA amplitude, 200-400μs pulse width, monopolar or bipolar configurations
Duration	Assessed throughout 4-week period	4 weeks (leads implanted), used during ~5.5 hours of motor activities

Outcomes

Outcome	Type	Control	Intervention	P-value
Safety - assessed by systematically documenting all adverse events	Primary	N/A	14 mild adverse events, 0 moderate or severe adverse events, 0 serious adverse events. All AEs resolved without sequelae
Assistive effects on strength (MVC across joints)	Secondary	Baseline MVC values	Mean +32% increase across joints (shoulder flexion +28%, elbow extension +35%, elbow flexion +34%, grip +55%)	Significant for multiple joints (bootstrap 95% CI)
Assistive effects on FMA motor score (week 2)	Secondary	SCS OFF: mean change from baseline +1.00	SCS ON: mean +5.57 points from baseline	Significant (bootstrap analysis)
Effective improvement on FMA motor score (week 4)	Secondary	Baseline FMA	SCS ON at week 4: mean +6.57 points from baseline	Significant
Therapeutic effects on FMA motor score (week 4 OFF)	Secondary	Baseline FMA OFF	Week 4 SCS OFF: mean +5.14 points from baseline	Significant
Assistive effects on arm kinematics (reach phase)	Secondary	SCS OFF condition	Significant improvements in path efficiency (+14.6%), log dimensionless jerk (+9.6%), velocity peaks (n.s.)	p<0.05 for most metrics
Therapeutic effects on spasticity (MAS total score)	Secondary	Baseline MAS	Week 4 OFF: mean reduction of 5.57 points across participants	Clinically significant (MCID >0.76)
Hand/finger function restoration	Secondary	Baseline hand function	3/7 participants with MEP+ or residual hand opening regained hand/finger movements with SCS ON	Individual participant analysis
Mild AE - Pain at lead exit site	Adverse	N/A	1 event (SCS01) - resolved with suture replacement
Mild AE - Shortness of breath	Adverse	N/A	1 event (SCS04) - related to stimulation at C3/C4 levels at high amplitude (5mA), resolved immediately by discontinuing stimulation. Protocol modified
Mild AE - Lead migration	Adverse	N/A	1 event (SCS05) - mediolateral migration, no negative effects on motor function
Other mild AEs	Adverse	N/A	11 additional mild AEs across participants (phlebitis, bruising, headache, vomiting, pain, etc.) - all resolved spontaneously or with minimal intervention

Subgroup Analysis

Exploratory analysis suggested spared sensory function (FMA sensory scores) may be a stronger determinant of responsiveness than baseline motor impairment or CST integrity. MEP+ status associated with hand/finger function restoration but not overall FMA improvement. Hemorrhagic stroke patients also benefited.

Criticisms

Small sample size (n=7) limits generalizability and statistical power
No control group performing only motor assessments - limits interpretation of therapeutic effects
Open-label design - participants and some assessors aware of stimulation status
Short intervention duration (4 weeks) - long-term effects unknown
Heterogeneous cohort with variable stroke types, locations, and severities
Protocol modification after adverse event (avoiding C3/C4 stimulation) may have limited proximal muscle recruitment
Minimal motor activity performed (8.6 hours total) - combination with formal rehabilitation not tested
Temporary implant only - long-term implantation feasibility not assessed
Single-center study
Lack of sham stimulation control
Evaluator blinding only for FMA assessments, not other outcomes

Funding

National Institutes of Health Brain Initiative grant no. UG3NS123135-01A1 and internal funding from University of Pittsburgh Department of Neurological Surgery, Carnegie Mellon University Department of Mechanical Engineering and Neuroscience Institute, and University of Pittsburgh Department of Physical Medicine and Rehabilitation

Based on: SCS for Post-Stroke Hemiparesis (medRxiv, 2025)