A HARD PILL TO SWALLOW: TRANSLATIONAL GENETICS OF PRESCRIPTION OPIOIDS FROM GENETIC TO CLINICAL MODELS

Overall Abstract Prescription opioid use remains a major pathway to opioid use disorder (OUD), yet the mechanisms that distinguish therapeutic exposure from escalation to problematic use remain poorly understood. This symposium brings together four complementary talks that examine how genetic data, longitudinal clinical models, and psychiatric comorbidity analyses can identify risk for prescription opioid-related outcomes and inform future translational efforts in psychiatry and pain medicine. Jean Gonzalez will present findings from the Prescription Opioid Genetics Study, based on 141,897 participants, showing that subjective effects experienced the first time prescription opioids strongly predict later OUD. Predictive models achieved high discrimination, and a brief two-question screener separated high- from low-risk individuals with robust performance. The first GWAS of subjective opioid effects identified genome-wide significant loci, including OPRM1, and showed substantial genetic overlap between positive subjective effects and OUD. Alexander Hatoum will present the Pain to Opioid use disorder Pattern through Sequence (POPS), a novel longitudinal phenotype derived from electronic health records using long short-term memory modeling in All of Us. POPS more strongly predicts OUD risk than chronic pain or long-term opioid prescription alone and captures genetic, behavioral, and environmental correlates of vulnerability, with genome-wide analyses identifying associated loci across ancestries. Andrea Georgiou will discuss the role of psychiatric comorbidities by presenting convergent evidence on the shared genetic and molecular architecture of OUD with major depressive disorder, schizophrenia, bipolar disorder, and post-traumatic stress disorder. Using genetic correlation, Mendelian randomization, transcriptome-wide association studies, and pathway enrichment analyses across brain tissues, this work identifies shared genes and implicates endocrine, immune, calcium signaling, and glutamatergic synapse pathways in OUD-related comorbidity. Ellen Tsai will close with findings from the largest GWAS to date of OUD and long-term opioid exposure in five EHR-linked biobanks spanning more than one million individuals across ancestries. By comparing case definitions based on OUD ICD codes alone versus ICD codes combined with long-term opioid prescribing, this work shows that broader case definitions can increase discovery power, while also highlighting the need to disentangle OUD-related liability from pain-related genetic signal. Abraham Palmer will be the discussant. Altogether, these talks will highlight how integrating genetic and clinical models can improve phenotypic resolution, identify mechanisms of vulnerability, and advance the long-term goal of earlier identification and more precise intervention for OUD.