Neuromodulation for post-stroke aphasia rehabilitation: established cortical stimulation and emerging network-based targets

Abstract

Post-stroke aphasia is a disabling disorder of language and communication for which speech-language therapy remains the standard of care. Because behavioural treatment yields variable gains and many people retain chronic communication limitations, non-invasive neuromodulation has been investigated to increase the responsiveness of residual language networks to therapy. This narrative review critically integrates evidence available through July 2026 for repetitive transcranial magnetic stimulation (rTMS), theta-burst stimulation, transcranial direct current stimulation (tDCS), and emerging approaches targeting the cerebellum, spinal sensorimotor pathways, and cellular metabolism. Recent meta-analyses support small-to-moderate average benefits of adjunctive cortical stimulation, especially for naming, but these estimates coexist with marked heterogeneity, small trials and incomplete evidence for generalisation, durability, functional communication, and quality of life. Recent randomised studies illustrate the value of pairing stimulation with a defined learning task and of using lesion, tract, functional-connectivity, or hemodynamic information to select targets and predict response. Cerebellar tDCS, transcutaneous spinal direct current stimulation, transcranial alternating current stimulation, and transcranial photobiomodulation are conceptually attractive but remain investigational because the clinical evidence is limited, inconsistent, or based on uncontrolled and proof-of-concept studies. Across modalities, serious adverse events are uncommon when contemporary screening and dosing guidance are followed, although stroke-specific comorbidity and seizure risk require trained supervision. Neuromodulation should therefore be viewed as a potentially useful, state-dependent amplifier of speech-language therapy rather than a replacement for it. Progress towards routine care will require adequately powered multisite trials, preregistered patient-level stratification, treatment-fidelity reporting, ecologically valid outcomes, longer follow-up, and direct comparisons of personalised with fixed protocols.