Aims:
The International Agency for Research on Cancer recently classified opium consumption as carcinogenic to humans, causing larynx, lung and bladder cancer. We evaluated the relationship between pharmaceutical opioid use and cancer incidence in a New South Wales (NSW) prospective cohort study.
Methods:
Cox proportional hazards regressions were used to calculate hazard ratios (HR) and 95% confidence intervals (CI) for cancer incidence in relation to opioid dispensing claims in the 12 months prior to baseline, among 64,464 (24%) of 267,357 participants in the 45 and Up Study (2005-2009). Opioid dispensing claims were ascertained via record linkage to the Pharmaceutical Benefits Scheme (PBS; 2005-2009; provided by Services Australia). Cancer incidence was ascertained via linkage to the NSW Cancer Registry to December 2019. Participants diagnosed with cancer prior to baseline, participants without concessional PBS claims, and participants who died within six months of baseline were excluded. Participants were censored at the end of the study period (December 2019), cancer diagnosis, or death (ascertained from the NSW Registry of Births, Deaths and Marriages to December 2019). Record linkage was performed by the Centre for Health Record Linkage (CHeReL). Secure data access was provided by the Secure Unified Research Environment (SURE). Exposure variables included number of opioid dispensing claims, annual per person oral morphine equivalents, opioid strength, duration of use (short-term vs. long-term), natural or synthetic opioids, opioid receptor type, and opioid acting duration. Regressions were adjusted for potential confounding factors including socio-demographic characteristics such as sex and age, and modifiable factors such as smoking and alcohol.
Results:
Of the included participants, 12,829 (19.1%) had at least one opioid dispensing claim within 12 months prior to baseline. Over a median 11.3 years follow-up, 12,277 (19%) participants were diagnosed with cancer. Participants with at least one opioid dispensing claim prior to baseline had significantly higher risk of lung cancer (HR=1.18; 95%CI=1.04-1.34), respiratory cancers combined (1.18; 1.04-1.33), and urinary cancers combined (1.22; 1.01-1.47) compared to participants who were not dispensed opioids. Risk for oesophageal, liver, pancreatic and bladder cancer, respiratory cancers combined and all cancers combined was significantly increased in association with the other exposure variables examined.
Discussion and Conclusion:
Pharmaceutical opioid use is associated with increased risk of oesophageal, liver, pancreatic, lung and bladder cancer, respiratory and urinary cancers combined, and all cancers combined. These data will be key to informing the Opioid Cohort Consortium (OPICO), a global pooled analysis of opioids and cancer risk.