Abstract
BACKGROUND: Short-term neurobiological effects of opioids are well characterized, yet the long-term consequences of substance exposure on malignancy risk remain poorly understood. Ethical constraints limit prospective analysis of opioid exposure and primary malignant brain tumor (PMBT) risk in human patients, making large-scale electronic health record (EHR) analyses essential. This study evaluates the 10-year incidence of PMBTs among adults with opioid-abuse disorder related diagnoses exposed to either buprenorphine or naltrexone. METHODS: We conducted a retrospective cohort study using the Epic Cosmos EHR database. Cohorts were defined by exclusive exposure to buprenorphine or naltrexone, with a non-exposed comparator cohort. PMBT incidence within 1-120 months post-exposure was identified. Due to platform constraints, multivariable regression could not be performed; instead, confounding was addressed using Mantel-Haenszel stratification across age, sex, and race. RESULTS: There was no significant difference in 10-year PMBT incidence between patients exposed exclusively to buprenorphine versus naltrexone (Mantel-Haenszel OR = 1.028, 95% CI 0.968-1.092; p = 0.381), however PMBT risk was lower for opioid disorder patients receiving either drug than for the control group. Sex-specific analyses suggested women exposed to naltrexone had significantly lower PMBT incidence relative to buprenorphine exposure (OR = 0.783, 95% CI 0.617-0.994; p = 0.044), whereas men showed a non-significant increase (OR = 1.211, 95% CI 0.939-1.564; p = 0.139). CONCLUSIONS: Buprenorphine and naltrexone were associated with comparable overall 10-year PMBT risk among adults with opioid-related diagnoses. The sex-specific interaction suggests naltrexone may be associated with reduced PMBT incidence in women, warranting further investigation. These findings contribute to the long-term neuro-oncologic safety profiles of opioid use disorder treatments.