OBJECTIVES: Older adults have the highest prevalence of non-cancer chronic pain conditions (NCPCs) that are commonly treated with prescription opioids. Animal studies show that opioids may not only increase the risk of developing cancer but also stage of cancer. This study investigated the association of the long-term prescription opioid use (LTPOU) with stage of cancer diagnosis among older cancer survivors with any NCPC.

METHODS: A retrospective cohort study design with data from the linked Surveillance, Epidemiology, and End Results and Medicare claims on older adults with NCPCs and cancer diagnoses (breast, prostate, colorectal non-Hodgkin lymphoma [NHL] (N=63,162) was used. LTPOU was defined as opioid use for at least 90 days prior to cancer diagnosis with an allowable 30-day gap between prescriptions. We allowed a washout period of 30 days to ensure that opioids were not prescribed for cancer symptoms. The outcome was stage of diagnosis: advanced stages(III and IV) and early stages of cancer(I and II). Adjusted logistic regression and selection-bias corrected regressions with inverse probability weights(IPW) controlled for biological, socioeconomic, behavioral, socio-cultural, medical-related, and access to care factors.

RESULTS: In our study cohort, 41.0%, 30.8%, 19.2%, and 9.0% had breast cancer, prostate, colorectal, and 9.0% NHL respectively. 74.7% had early stage and 18.1% had advanced stage cancers. Advanced cancer patients had 11.6% LTPOU and early-stage cancer patients had 10.7% LTPOU. Adjusted regression models revealed that LTPOU was significantly associated with higher odds of advanced stage of cancer compared to the no opioid use group (AOR: 1.10, 95%CI= 1.02,1.19]. Accounting for selection bias with IPW did not alter the findings.

CONCLUSIONS: One in nine adults with cancer had LTPOU before cancer diagnosis. Those with LTPOU were more likely to be diagnosed with advanced cancer compared to those without LTPOU. These findings need to be confirmed with well-designed prospective studies.