Spinal cord stimulation in the setting of oncologic pain: a systematic review

Abstract

Background: Cancer-related pain remains a major clinical challenge, particularly when neuropathic or refractory to pharmacologic and interventional measures. Spinal cord stimulation (SCS) has been increasingly recognized as a potential therapy for select patients with cancer-related pain. Objective: This study aims to systematically review the evidence on the effectiveness, safety, and clinical applicability of SCS for cancer-related pain. Methods: A structured search of PubMed, MEDLINE, Embase, and Cochrane databases (inception to 2025) was performed for studies evaluating SCS for cancer-related pain. Eligible study types included randomized controlled trials (RCTs), observational studies, prospective/retrospective cohorts, case series, and case reports. Outcomes included pain relief, functional improvement, opioid reduction, durability, and complications. Results: Thirty-four studies met inclusion criteria (0 RCTs, 5 cohort studies, 29 case series/reports), comprising 204 patients. Most studies involved neuropathic pain due to tumor infiltration, post-surgical nerve injury, radiation-induced neuropathy, or chemotherapy-induced peripheral neuropathy. Across these studies it was found that 60-80% of patients achieved ≥50% pain relief during trial stimulation and 50-70% maintained benefit at follow-up (weeks to years). Opioid reduction was reported in 30-60% of cases. Complication rates were comparable to non-cancer SCS populations but theoretically increased by immunosuppression and coagulopathy. However, evidence quality was uniformly low due to publication bias, small samples, and heterogeneity. Conclusion: SCS appears to provide meaningful pain relief for select patients with refractory cancer-related neuropathic pain; however, the evidence is limited to small observational studies, with no RCTs. Multidisciplinary selection and individualized risk-benefit assessment remain crucial. High-quality comparative trials and prospective registries are needed.