Neurophysiological and Psychophysical Biomarkers for Predicting Spinal Cord Stimulation Treatment Response in Chronic Neuropathic Pain: A Systematic Review

Abstract

Background: Spinal cord stimulation (SCS) is an established therapy for chronic neuropathic pain, yet a substantial minority of implanted patients do not obtain durable relief, and patient selection still relies largely on a subjective percutaneous stimulation trial. Objective neurophysiological and psychophysical biomarkers could enable precision selection, but their predictive value has not been consolidated across modalities. Objectives: To evaluate whether pre-treatment or intraoperative neurophysiological (electroencephalography (EEG), magnetoencephalography (MEG), somatosensory evoked potentials (SSEP), evoked compound action potentials (ECAP)) and psychophysical (conditioned pain modulation (CPM), quantitative sensory testing [QST]) biomarkers predict SCS treatment response in adults with chronic neuropathic pain. Methods: Following a pre-registered protocol (International Prospective Register of Systematic Reviews (PROSPERO CRD420261410302)) and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guideline, we searched six databases and two registries; the final search was run on 4 June 2026. Records underwent dual, independent screening in Rayyan; data were extracted in duplicate and risk of bias was assessed with the Quality In Prognosis Studies (QUIPS) tool, the Prediction model Risk Of Bias Assessment Tool (PROBAST), or the Quality Assessment of Prognostic Accuracy Studies (QUAPAS) tool. Because of clinical and methodological heterogeneity, findings were synthesized narratively using the Synthesis Without Meta-analysis (SWiM) guideline. Results: Twelve peer-reviewed studies met eligibility. Most were prognostic-factor or responder-discrimination studies rather than prospectively validated predictors; two were multivariable machine learning prediction models and one used ECAP-derived neural dose as a therapy-embedded surrogate. The strongest single-study signal was the preoperative SSEP (one 2003 cohort, n = 95), in which normal dorsal-column central conduction was associated with a 75% success rate; this legacy finding requires replication with contemporary SCS paradigms. Resting EEG/MEG spectral features (alpha, gamma) discriminated responders in several small cohorts. The two prediction models reported apparent internal accuracy of 76–88% (area under the curve up to 0.88) but neither underwent external validation. QST and CPM findings were promising but directionally inconsistent. All 12 studies were at high overall risk of bias. Conclusions: Objective biomarkers—most notably the preoperative SSEP and resting-state EEG/MEG spectral features—show promise for predicting SCS response, but the evidence is preliminary, heterogeneous, small, and lacks external validation. Adequately powered, prospective, externally validated studies are needed before clinical adoption.

The authors' abstract, as published at the source. Journal of Personalized Medicine, 2026 · DOI ↗