894 Comparison of Intraoperative Neural Changes During Tonic Versus High-Frequency Spinal Cord Stimulation in Chronic Pain Patients

Abstract

INTRODUCTION: Chronic pain is a prevalent issue that may affect up to 34% of US adults, and spinal cord stimulation (SCS) is a promising therapy. However, SCS device selection remains a challenge because there are currently no objective biomarkers to evaluate chronic pain, or to predict which patients will be responders. METHODS: Intraoperative EEG signals were recorded in 24 chronic pain patients (14 women) during permanent SCS electrode placement surgery. Neural responses were captured with respect to tonic (60 Hz/300 µs) and high-frequency (1 kHz/30 µs) SCS paradigms in a randomized manner and correlated to Numeric Rating Scale (NRS) scores. All data were analyzed offline in MATLAB and SPSS v28.0.1.1. RESULTS: Our analyses for the main effects demonstrated a significant difference between cortical regions (F(2.88,57.7) = 7.60 , p < 0.001) and a marginal difference between stimulation conditions (F(2,40) = 2.970, p = 0.063). The post hoc analysis for these factors revealed that alpha/theta peak power ratio was significantly different between in central regions (C3-C4) between tonic and HF stimulation (p = 0.006) and temporal regions (T3-T4) between baseline and HF stimulation (p = 0.049). The comparison of NRS scores before and 3 months after surgery indicated a significant improvement in pain intensity (p < 0.001) at the group level. The neural features found to be significantly correlated with preoperative NRS scores in PFC (p = 0.036 and p = 0.025) and temporal regions (p = 0.041 and p = 0.012) under both stimulations. CONCLUSIONS: Our preliminary results support the possibility that different stimulation patterns affect the electrophysiological activity differently in various cortical regions. Such differences might account for the variability in pain suppression in patients with chronic pain treated with SCS. Future work will utilize this dataset to develop a machine learning classifier to support objective device selection strategies.