UC San Diego

BACKGROUND: Cancer treatment is often complicated by co-occurring substance dependence such as alcohol, cannabis, opioids, amphetamines, and tobacco. The objective of this study was to analyze the association between substance use (alcohol, cannabis, opioids, amphetamines, and tobacco) and mortality among cancer patients. METHODS: This retrospective study used de-identified data from the University of California, San Diego Health on 22,763 cancer patients aged ≥ 18 years. Substance dependance post-diagnosis was identified based on clinical observations from electronic health records. Tumor staging was determined using the TNM system, with missing data imputed via biomarker-based regression. Cox proportional hazards models were computed to assess severity-adjusted associations between substance dependance, by type, and mortality. RESULTS: Cannabis dependence was not significantly associated with mortality. Alcohol, tobacco, opioid, and amphetamine dependence were linked to increased mortality in both unadjusted and adjusted models. Adjusting for tumor stage reduced hazard ratios for opioid and tobacco dependence, suggesting stage as a confounding factor. Alcohol and opioid dependence remained significant predictors of mortality, while findings for cannabis and amphetamine were limited by small sample sizes. CONCLUSION: Alcohol and opioid dependence in cancer patients was observed to be a significant predictor of mortality after adjusting for disease severity, emphasizing the need for targeted interventions among this subpopulation.