BACKGROUND: Epidemiologic studies have shown an inverse association of Parkinson's disease (PD) and cancer occurrence overall, but a positive association with malignant melanoma. Whether the increased risk of melanoma is due to the disease or its treatment is unknown. In preparation for a safety study of antiparkinsonian drugs (APD) and melanoma in US Medicare claims data, we conducted a feasibility assessment to evaluate potential study size and follow-up time.
OBJECTIVES: To describe the characteristics of PD patients initiating APD, including uptake and utilization of rasagiline, a monoamine oxidase-B inhibitor approved in the US in 2006, and follow-up time.
METHODS: This was a descriptive retrospective cohort analysis of US Medicare claims data (2006-2011). The sample comprised individuals aged ≥ 65 years with fee-for-service Medicare Parts A, B and D insurance, with ≥ 2 outpatient/physician claims for PD (ICD-9 code: 332.0) or ≥ 1 inpatient claim for PD, and with first exposure to either (1) rasagiline or (2) another APD after at least 6-months of enrollment. Characteristics of the study cohorts were compared.
RESULTS: Compared with the other APD initiator cohort (n=120,262), the rasagiline initiator cohort (n=14,170) included on average a slightly younger population (76 vs. 78 years), a higher proportion of males (54% vs. 47%) and a lower proportion of individuals receiving a Medicare low-income subsidy (25% vs. 44%). At cohort entry, 91% of rasagiline initiators and 57% of other APD initiators had prior use of an APD other than the qualifying drug. The proportion of patients with ≥ 3 years of follow-up in each cohort was similar (28% to 27%). Annual losses to follow-up occurred in 10%-11% of each cohort, mainly because of death.
CONCLUSIONS: With the observed uptake of rasagiline in US seniors, an estimated additional 2 years of data may be necessary before accrual of sufficient follow-up time to evaluate melanoma as an outcome and conduct comparative exposure analyses. Initial results suggest that rasagiline initiators may differ from other APD initiators and thus could affect the relative risk of melanoma.
