Personalized, closed-loop deep brain stimulation for chronic pain

Abstract Background Chronic pain is a major healthcare problem associated with maladaptive brain circuit changes - many patients are unresponsive to all available therapies. Deep brain stimulation (DBS) is a promising treatment, but traditional targets have not proven consistently effective. DBS for pain may be improved by individualizing location and timing of stimulation based on real time brain measurements and patient reports. Methods To optimize personalized stimulation targets, we first performed a double blind, sham-controlled, intracranial EEG brain mapping trial spanning 10 hospital days, in six participants with refractory neuropathic pain syndromes. Five participants with clinically meaningful pain relief were then implanted with permanent devices capable of brain stimulation and recording. We used ambulatory electrical brain recordings to derive bespoke pain biomarkers using machine learning. Pain biomarkers were used in closedloop DBS algorithms for personalized therapy. After an open-label period, we tested the feasibility and efficacy of closed-loop DBS for pain relief against sham in a double-blind, cross-over trial. Results Trial intracranial testing revealed multiple brain targets among cortico-striatal-thalamocortical pathways that produced rapid pain relief across participants. We predicted individual pain metrics from both inpatient and ambulatory brain activity with high accuracy; pain biomarkers were incorporated into closed-loop DBS algorithms that also responded to sleep-wake cycles. Personalized, closed-loop DBS was superior to sham, with durability up to 3.5 years. Conclusions Precision-medicine DBS, with individually optimized brain stimulation targets and closed-loop delivery of stimulation in response to pain biomarkers, is a feasible strategy to treat refractory chronic pain syndromes. Trial Registration NCT04144972