VNS-REHAB 1-Year Outcomes
Long-Term Outcomes of Vagus Nerve Stimulation Paired With Upper Extremity Rehabilitation After Stroke
Year of Publication: 2025
Authors: Teresa J. Kimberley, PhD; Steven C. Cramer, MD; Steven L. Wolf, ..., MD; VNS-REHAB Trial Group
Journal: Stroke
Citation: Stroke. 2025 Aug;56(8):2255-2265. doi: 10.1161/STROKEAHA.124.050479
Link: https://www.ahajournals.org/doi/10.1161/STROKEAHA.124.050479
PDF: https://www.ahajournals.org/doi/epub/10....EAHA.124.050479
PDF: Clinical Question
What are the long-term effects (through 12 months) of vagus nerve stimulation (VNS) paired with rehabilitation on impairment, activity, and participation in people with upper extremity (UE) impairment after chronic ischemic stroke?
Study Overview
Objective
To assess the long-term safety and durability of functional improvements from vagus nerve stimulation (VNS) paired with rehabilitation in patients with chronic stroke-related upper limb weakness.
Study Summary
At 1-year follow-up, VNS paired with rehab remained safe and showed sustained improvement in upper limb function. Most participants who responded at day 90 continued to benefit at 12 months.
Intervention
Patients previously enrolled in the VNS-REHAB trial (randomized to VNS + rehab or sham + rehab) were followed for one year post-treatment. After the primary endpoint (day 90), sham group participants were offered the opportunity to receive active VNS implantation and therapy.
Patients per Arm
Initially: VNS: 53, Sham: 55. Of the sham group, 45 chose to receive VNS post-trial.
Bottom Line
Vagus nerve stimulation (VNS) paired with rehabilitation in individuals with chronic UE deficits after ischemic stroke resulted in maintained improvements in UE impairment, activity, participation, and quality-of-life measures at 1 year. This approach is a beneficial treatment option for long-term recovery.
Major Points
- This is a post hoc analysis of the VNS-REHAB randomized clinical trial, reporting unblinded, partial crossover, and pooled 1-year outcomes. Initially, 108 participants were enrolled.
- Participants received 6 weeks of in-clinic intensive task-specific rehabilitation and 3 months of home-based exercise with active or sham VNS. Control participants then crossed over to receive active stimulation.
- 74 participants (69%) completed 1-year follow-up and provided pooled data.
- At 1 year, compared with baseline, there were significant improvements in impairment (Fugl-Meyer Assessment UE [FMA-UE], 5.23 points [95% CI, 4.08-6.39]; P<0.001) and activity (Wolf Motor Function Test [WMFT], 0.50 points [95% CI, 0.41-0.59]; P<0.001).
- Significant improvements were also seen in patient-reported outcomes (Motor Activity Log-Quality of Movement: 0.64 [95% CI, 0.46-0.82], P<0.001; Motor Activity Log-Amount of Use: 0.64 [95% CI, 0.46-0.82], P<0.001; Stroke Impact Scale-Activities of Daily Living: 7.43 [95% CI, 5.09-9.77], P<0.001; Stroke Impact Scale-Hand: 17.89 [95% CI, 14.16-21.63], P<0.001; EQ-5D: 5.76 [95% CI, 2.08-9.45], P<0.05; and Stroke Specific-Quality of Life: 0.29 [95% CI, 0.19-0.39], P<0.001).
- 66.2% of pooled participants (49 of 74) achieved at least a minimal clinically important difference (MCID) improvement in FMA-UE (≥6 points) or WMFT (≥0.4 points) at 1 year.
- No serious adverse events related to therapy or stimulation were reported in the long-term phase. No serious adverse events led to device explantation.
Design
Study Type: Post hoc analysis of a pivotal randomized clinical trial (VNS-REHAB)
Randomization: 1
Blinding: Initially blinded (6 weeks in-clinic + 3 months home exercise); unblinded after Day-90 outcome assessment for crossover and long-term phases.
Enrollment Period: Original VNS-REHAB trial: October 2, 2017 to September 12, 2019
Follow-up Duration: 1 year (after completion of in-clinic therapy)
Centers: 19
Countries: United States, United Kingdom
Sample Size: 74
Analysis: Pooled data analysis from both randomized groups (as all received active paired VNS for similar periods). Linear mixed-effects regression model with random-intercept model for FMA-UE, WMFT, and PROMs, controlling for side of paresis, sex, time since stroke, and age. Bonferroni adjustment for long-term PROM P-values. RStudio (version 2023.06.1). Response profiles calculated based on MCID.
Inclusion Criteria
- Participants aged between 22 and 80 years with a history of unilateral supratentorial ischemic stroke occurring 9 months to 10 years before enrollment.
- Moderate-to-severe arm impairment, defined as an FMA-UE score between 20 and 50.
- At least minimal active wrist flexion/extension and active abduction/extension of the thumb and at least 2 additional digits.
- Right- or left-sided upper extremity weakness (per NCT03131960).
- Ability to communicate and follow two-step commands (per NCT03131960).
- Full inclusion criteria are detailed in the Supplemental Material and the trial registry (NCT03131960).
Exclusion Criteria
- Full exclusion criteria are detailed in the Supplemental Material of the paper and in the clinicaltrials.gov registry entry for NCT03131960 (which enumerates 14 exclusion criteria); specific exclusions are not itemized within the main text of this 1-year outcomes paper.
Baseline Characteristics
Control:
- Sex, n (%) - Male: 26 (72%)
- Sex, n (%) - Female: 10 (28%)
- Age, y; mean±SD: 60.6 (8.4)
- Race, n (%) - Black: 6 (17%)
- Race, n (%) - Asian, Indian, or other: 2 (6%)
- Race, n (%) - White: 27 (75%)
- Race, n (%) - Not reported: 1 (3%)
- Dominant hand, n (%) - Right: 32 (89%)
- Dominant hand, n (%) - Left: 4 (11%)
- Dominant hand, n (%) - Ambidextrous: 0 (0%)
- Side of stroke, n (%) - Left: 19 (53%)
- Side of stroke, n (%) - Right: 17 (47%)
- Time since stroke, y; mean (SD): 3.6 (2.8)
- Baseline FMA-UE, mean (SD): 35.6 (8.6)
- Baseline WMFT-FAS, mean (SD): 2.8 (0.7)
Active:
- Sex, n (%) - Male: 25 (66%)
- Sex, n (%) - Female: 13 (34%)
- Age, y; mean±SD: 58.8 (9.5)
- Race, n (%) - Black: 8 (21%)
- Race, n (%) - Asian, Indian, or other: 1 (3%)
- Race, n (%) - White: 28 (74%)
- Race, n (%) - Not reported: 1 (3%)
- Dominant hand, n (%) - Right: 33 (87%)
- Dominant hand, n (%) - Left: 4 (11%)
- Dominant hand, n (%) - Ambidextrous: 1 (3%)
- Side of stroke, n (%) - Left: 21 (55%)
- Side of stroke, n (%) - Right: 17 (45%)
- Time since stroke, y; mean (SD): 3.0 (2.0)
- Baseline FMA-UE, mean (SD): 34.2 (8.1)
- Baseline WMFT-FAS, mean (SD): 2.7 (0.7)
Pooled (active and control):
- Sex, n (%) - Male: 51 (69%)
- Sex, n (%) - Female: 23 (31%)
- Age, y; mean±SD: 59.6 (8.9)
- Race, n (%) - White: 55 (74%)
- Race, n (%) - Black: 14 (19%)
- Side of stroke, n (%) - Left: 40 (54%)
- Side of stroke, n (%) - Right: 34 (46%)
- Time since stroke, y; mean (SD): 3.3 (2.5)
- Baseline FMA-UE, mean (SD): 34.9 (8.3)
- Baseline WMFT-FAS, mean (SD): 2.73 (0.7)
Arms
| Field | Active VNS Group (Original Randomization) | Control |
|---|---|---|
| Intervention | VNS device implantation followed by 18 sessions of in-clinic intensive task-specific rehabilitation paired with active VNS. Then, 3 months of self-initiated home-based exercise with active VNS (activated via magnet swipe). Followed by 1 year of home-based exercises with self-initiated active VNS. | VNS device implantation followed by 18 sessions of in-clinic intensive task-specific rehabilitation paired with sham VNS. Then, 3 months of self-initiated home-based exercise with sham VNS. After unblinding, crossed over to receive another 18 sessions of in-clinic rehabilitation with active VNS, followed by 1 year of home-based exercises with self-initiated active VNS. |
| Duration | 6 weeks in-clinic + 3 months home + 1 year long-term | 6 weeks in-clinic (sham) + 3 months home (sham) + 6 weeks in-clinic (active) + 1 year long-term (active) |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Co-primary outcomes for the 1-year long-term analysis (per Methods): change from baseline to 12 months in FMA-UE (impairment) and WMFT (activity), analyzed with a linear mixed-effects model in pooled participants who received active paired VNS (n=74; original active group + crossover controls). | Primary | FMA-UE: 5.23 points (95% CI, 4.08-6.39); WMFT: 0.50 points (95% CI, 0.41-0.59) | <0.001 (both) | ||
| Motor Activity Log-Quality of Movement change from baseline to 1 year (pooled) | Secondary | 0.64 (95% CI, 0.46-0.82) | <0.001 | ||
| Motor Activity Log-Amount of Use change from baseline to 1 year (pooled) | Secondary | 0.64 (95% CI, 0.46-0.82) | <0.001 | ||
| Stroke Impact Scale-Activities of Daily Living change from baseline to 1 year (pooled) | Secondary | 7.43 (95% CI, 5.09-9.77) | <0.001 | ||
| Stroke Impact Scale-Hand change from baseline to 1 year (pooled) | Secondary | 17.89 (95% CI, 14.16-21.63) | <0.001 | ||
| EQ-5D change from baseline to 1 year (pooled) | Secondary | 5.76 (95% CI, 2.08-9.45) | <0.05 | ||
| Stroke Specific-Quality of Life change from baseline to 1 year (pooled) | Secondary | 0.29 (95% CI, 0.19-0.39) | <0.001 | ||
| Beck Depression Inventory change from baseline to 1 year (pooled) | Secondary | No significant change | NS | ||
| FMA-UE change from baseline to Day-1_cross (control group crossover to active VNS) | Secondary | 6.13 (95% CI, 4.78-7.48) | <0.001 | ||
| FMA-UE change from baseline to Day-90_cross (control group crossover to active VNS) | Secondary | 5.64 (95% CI, 4.22-7.06) | <0.001 | ||
| WMFT change from baseline to Day-1_cross (control group crossover to active VNS) | Secondary | 0.42 (95% CI, 0.32-0.53) | <0.001 | ||
| WMFT change from baseline to Day-90_cross (control group crossover to active VNS) | Secondary | 0.41 (95% CI, 0.29-0.52) | <0.001 | ||
| Response profile: proportion achieving MCID on FMA-UE (≥6 pts) and/or WMFT (≥0.4 pts) at 1 year | Secondary | 49/74 (66.2%) | |||
| Serious adverse events in long-term phase related to therapy or stimulation | Adverse | 0 | 0 | ||
| Serious adverse events after blinded phase and before 1-year visit (all unrelated to intervention) | Adverse | 1 | 3 | ||
| Device explantation attributable to a serious adverse event | Adverse | 0 | 0 |
Criticisms
- This is a post hoc analysis of unblinded, partial crossover, and pooled data, which limits the ability to draw definitive conclusions about between-group differences in the long-term phase due to the lack of a control comparator.
- The study has missing data, particularly for longer-term follow-up visits, though this was attributed to pandemic-related restrictions and deemed 'missing at random'.
- The study did not include specific steps to ensure the diversity of the sample or the steering committee, potentially limiting generalizability beyond the enrolled population.
- The self-initiated home exercise was not explicitly monitored, which precludes definitive conclusions related to the specific rehabilitation components that may have contributed to maintaining participants' gains or including adherence to models to explain variance over time.
Subgroup Analysis
Side of stroke, sex, age, and time since stroke did not significantly impact FMA-UE or WMFT outcomes in the pooled analysis. No formal subgroup analyses were presented in the abstract for long-term outcomes.
Funding
MicroTransponder, Inc.
Based on: VNS-REHAB 1-Year Outcomes (Stroke, 2025)
Authors: Teresa J. Kimberley, PhD; Steven C. Cramer, MD; Steven L. Wolf, ..., MD; VNS-REHAB Trial Group
Citation: Stroke. 2025 Aug;56(8):2255-2265. doi: 10.1161/STROKEAHA.124.050479
Content summarized and formatted by NeuroTrials.ai.