BACKGROUND: Rasagiline is a selective, irreversible monoamine oxidase type B inhibitor used to treat Parkinson’s disease (PD). Several malignant melanoma cases were diagnosed during the rasagiline development program, prompting regulatory requests for a study to evaluate melanoma risk with rasagiline use.
OBJECTIVES: To estimate and compare the risk of cutaneous melanoma between initiators of rasagiline or other antiparkinsonian drugs (APD) in a PD population and evaluate possible surveillance bias due to potential increased skin cancer surveillance among rasagiline new users.
METHODS: This was a retrospective cohort study in the US Medicare claims research database. The study population comprised adults aged ≥65 years with diagnosis claims for PD and continuous enrollment for ≥6 months in Medicare Parts A, B, and D fee-for-service coverage who started rasagiline (Cohort A) or a non-rasagiline APD (Cohort B) in 2006-2015. Cohort B was matched to Cohort A (4:1) on age, sex, and index medication date. Incident melanoma cases were identified by an algorithm and validated by medical record review (positive predictive value 83%). Health care utilization before and after cohort entry and the incidence of nonmelanoma skin cancer during follow-up were compared between Cohorts A and B to evaluate surveillance bias.
RESULTS: There were 23,708 patients in Cohort A and 96,552 in Cohort B. The crude incidence rate of algorithm-defined melanoma cases per 100,000 person-years during the study period was 334 (95% confidence interval, 292-382) in Cohort A and 208 (190-228) in Cohort B (unadjusted rate difference 126 per 100,000 person-years (99% CI, 78-174). The adjusted Cox regression hazard ratio comparing Cohort A with Cohort B was 1.37 (1.14-1.65) and increased with longer duration of exposure and higher cumulative rasagiline dose. Patients in Cohort A were more likely than patients in Cohort B to have ≥ 1 dermatologist visit (15% vs. 11%) or ≥ 1 skin biopsy (8% vs. 6%) in the 180 days before cohort entry. A higher percentage of melanoma cases in Cohort A than Cohort B were early stage (63% vs. 58%). The crude incidence rate ratio of nonmelanoma skin cancer comparing Cohort A with B was 1.44 (1.35-1.54).
CONCLUSIONS: We found a small increased incidence of melanoma with exposure to rasagiline compared with other APD. Surveillance bias from increased skin cancer surveillance among new users of rasagiline is a possible contributing explanation for the observed results.
