Background and purpose: A reliable neuroimaging biomarker to predict language improvement after neuromodulation in post-stroke aphasia is lacking. It is hypothesized that aphasic patients with stroke injuries in the left primary language circuits but with sufficient right arcuate fasciculus (AF) integrity might respond to low-frequency repetitive transcranial magnetic stimulation (LF-rTMS), leading to language improvement. This study aimed to assess the microstructural indices of the right AF before LF-rTMS treatment and further correlate with language improvement after the treatment.
Methods: Thirty-three patients with at least 3 months after stroke in the left hemisphere and nonfluent aphasia were recruited in this randomized double-blind study. All patients received real 1-Hz LF-rTMS (n = 16) or sham stimulation (n = 17) at the right pars triangularis for 10 consecutive weekdays. Fractional anisotropy, axial diffusivity, radial diffusivity and apparent diffusion coefficient of the right AF were extracted using diffusion tensor imaging before the rTMS treatment and correlated with the measured functional improvement by the Concise Chinese Aphasia Test.
Results: The Concise Chinese Aphasia Test change scores revealed a stronger language improvement in auditory/reading comprehension and expression in the rTMS group than in the sham group. Regression analysis showed that the pre-treatment fractional anisotropy, axial diffusivity and apparent diffusion coefficient of the right AF significantly correlated with the expression abilities (R2 > 0.700, p < 0.044) and comprehension abilities (R2 > 0.702, p < 0.039) in the rTMS group.
Conclusions: It was concluded that the right AF could be a predictor in language recovery induced by LF-rTMS after the injuries of primary language circuits.
Keywords: aphasia; diffusion tensor image; linear regression; repetitive transcranial magnetic stimulation; stroke.
© 2023 European Academy of Neurology.