Patterns of Prescription Medication Use Before Diagnosis of Early Age-Onset Colorectal Cancer: Population-Based Descriptive Study

Abstract

Background Colorectal cancer (CRC) is estimated to be the fourth most common cancer diagnosis in Canada (except for nonmelanoma skin cancers) and the second and third leading cause of cancer-related death in male and female individuals, respectively. Objective The rising incidence of early age-onset colorectal cancer (EAO-CRC; diagnosis at less than 50 years) calls for a better understanding of patients’ pathway to diagnosis. Therefore, we evaluated patterns of prescription medication use before EAO-CRC diagnosis. Methods We used linked administrative health databases in British Columbia (BC), Canada, to identify individuals diagnosed with EAO-CRC between January 1, 2010, and December 31, 2016 (hereinafter referred to as “cases”), along with cancer-free controls (1:10), matched by age and sex. We identified all prescriptions dispensed from community pharmacies during the year prior to diagnosis and used the Anatomical Therapeutic Chemical Classification system Level 3 to group prescriptions according to the drug class. A parallel assessment was conducted for individuals diagnosed with average age-onset CRC (diagnosis at age 50 years and older). Results We included 1001 EAO-CRC cases (n=450, 45% female participants; mean 41.0, SD 6.1 years), and 12,989 prescriptions were filled in the year before diagnosis by 797 (79.7%) individuals. Top-filled drugs were antidepressants (first; n=1698, 13.1%). Drugs for peptic ulcer disease and gastroesophageal reflux disease (third; n=795, 6.1%) were more likely filled by EAO-CRC cases than controls (odds ratio [OR] 1.4, 95% CI 1.2-1.7) and with more frequent fills (OR 1.8, 95% CI 1.7-1.9). We noted similar patterns for topical agents for hemorrhoids and anal fissures, which were more likely filled by EAO-CRC cases than controls (OR 7.4, 95% CI 5.8-9.4) and with more frequent fills (OR 15.6, 95% CI 13.1-18.6). Conclusions We observed frequent prescription medication use in the year before diagnosis of EAO-CRC, including for drugs to treat commonly reported symptoms of EAO-CRC.