Abstract 104: Navigated Brain Stimulation for Upper Limb Recovery After Stroke: A Randomized Sham Controlled Clinical Trial of Low Frequency r-TMS to Non-injured Hemisphere Combined With Upper Limb Rehabilitation

Introduction: Upper limb function following stroke is limited, with only 50% regaining some function and less than 20% regaining normal function. Repetitive TMS has promise as an adjunct to upper limb therapy after stroke. We aimed to determine if navigated brain stimulation (NBS) with a low-frequency (1 Hz) protocol to non-injured hemisphere combined with upper limb rehabilitation would improve arm motor function better than rehabilitation alone. Method: We enrolled 199 patients with hemiplegia from ischemic or hemorrhagic stroke within 3-12 m post ictus. Randomization was in a 2:1 ratio to NBS with 18 sessions of rehabilitation over 6 weeks, or to sham NBS with therapy. The primary end-point was rate of achieving a 5-point improvement on the upper limb Fugl-Myer (ULFM) score at 6 m post-treatment and safety. Secondary outcomes included post-treatment, 1 m, 3 m and 6 m change on ULFM, action research arm test (ARAT), and EQ-5D-3L health questionnaire. Results: Of 199 subjects enrolled, 167 completed treatment and follow-up due to early stoppage of data collection after interim futility analysis. All subjects improved significantly on each outcome measure at each point of follow up, including 6 m post-treatment ULFM: treatment (8.1±7.4, p<0.001) and sham (8.5±8.7, p<0.001). In the ITT analysis, there was no difference on achievement of 5 points on ULFM at 6 m post-treatment: 67% treatment vs. 65% sham (chi-sq 1.105, p=0.76). Repeated measures ANCOVA group*time interaction showed no significant difference between groups for ULFM (p=0.87), ARAT (p=0.80) and the EQ 5D (p=0.96). There were no study or device related serious adverse events and no difference in SAE’s between groups. Conclusion: NBS can be safely used in the clinical setting. Clinically important gains were observed in both study arms suggesting no additional effect of 1 Hz NBS in stroke subjects within 3-12 m post ictus. The lack of NBS effects may be explained by the large effect size (ceiling effect) or potentially that the sham coil unintentionally induced cortical currents. Further analysis of the sham method and also secondary subgroup analyses will provide further insights and generate novel hypothesis to be confirmed in future NBS trials.