With the continuous advancements in critical medical care and neuroimaging technologies, there has been a significant improvement in the survival rate of patients who have suffered from brain injuries, including conditions such as cerebrovascular disease, stroke, anoxia, and traumatic brain injury (TBI).¹ Among them, a considerable number of severe cases have been diagnosed with disorders of consciousness (DOCs). DOCs are characterized by alterations in wakefulness and/or awareness,² mainly including coma, defined by the complete absence of wakefulness and awareness; unresponsive wakefulness syndrome (UWS), previously known as vegetative state (VS), which is characterized by the presence of wakefulness but without awareness of themselves and the external environment; and minimally conscious state (MCS), characterized by minimal, inconsistent, yet reproducible signs of awareness.^1,3 The accurate and prompt diagnosis of DOC is imperative for patients as it directly influences the design of individualized therapeutic strategies, the determination of prognostic outcomes, and even the decision-making process regarding end-of-life care.⁴

Deep brain stimulation (DBS) and spinal cord stimulation (SCS) have been investigated for over four decades as neuromodulatory treatments for chronic disorders of consciousness (DOC), including the vegetative state/unresponsive wakefulness syndrome (VS/UWS) and the minimally conscious state (MCS), for which no disease-modifying pharmacological therapy exists. This narrative review synthesizes literature published between 2016 and 2026 on the neurobiological rationale, clinical evidence, neurophysiological biomarkers, and safety of DBS and SCS for chronic DOC. The ‘mesocircuit hypothesis’, supported by human intraoperative electrophysiology and nonhuman primate causal manipulation, provides a coherent rationale for central thalamic and centromedian-parafascicular targeting. Across case series, cohort studies, and individual patient data meta-analyses, both DBS and SCS produce clinically meaningful improvement in a minority of patients, consistently more often in MCS (63–74% response) than in VS/UWS (26–30% response), with younger age and shorter injury-to-treatment interval predicting better outcomes. However, the two largest individual patient data meta-analyses of DBS concur that current evidence is insufficient to conclude that stimulation alters the natural history of neurological recovery beyond what would be expected from patient selection and spontaneous recovery, and adverse event rates (8.1% for SCS, 18.2% for DBS) indicate non-trivial surgical risk in this vulnerable population. Neurophysiological and neuroimaging biomarkers derived from EEG, functional near-infrared spectroscopy, functional MRI, and FDG-PET show promise for objective outcome prediction but await prospective validation. We conclude that DBS and SCS remain experimental, mechanistically motivated but not yet definitively proven therapies for chronic DOC, and that adequately powered randomized controlled trials, biomarker-guided patient selection, and dedicated ethical frameworks are needed before either modality can be recommended outside of research settings.