BACKGROUND: Postmarketing safety studies aimed at identifying whether a medication is potentially associated with a rare cancer are challenging to implement, especially without a national cancer registry as a single centralized data source.
OBJECTIVES: To describe innovative efforts undertaken with state cancer registries to assess a possible association between teriparatide treatment and osteosarcoma in humans and to provide an update on study progress.
METHODS: Two studies are underway: 1) a retrospective 15-year case series surveillance study, initiated in 2003 after initial drug approval; 2) a prospective 12-year patient registry linkage study, initiated in 2009 after a new indication was approved. In the retrospective study, incident cases of adult osteosarcoma diagnosed January 1, 2003, or later are identified through cancer registries, and exposure to possible risk factors is ascertained by telephone interview in the United States (US) and chart abstraction in five Nordic countries. In the prospective linkage study, patients enrolled in the voluntary US Forteo Patient Registry are linked annually to adult cases of osteosarcoma diagnosed January 1, 2009, or later in participating US state cancer registry databases.
RESULTS: In the retrospective study, as of December 31, 2012, 1,785 cases of adult osteosarcoma have been identified from 18 cancer registries for diagnosis years 2003- 2010. Patient/proxy interviews were completed for 24% of cases. In the linkage study, as of December 31, 2012, 30,758 patients had been registered, and the third annual linkage had been completed with 38 participating state cancer registries covering 86% of the US population aged 18 years and older. At this time, the studies do not support a causal association between teriparatide treatment and osteosarcoma in humans; however, both studies are still underway.
CONCLUSIONS: To monitor for a potential signal of a rare cancer event, it is necessary to apply innovative approaches to study design. It is also necessary to develop a study protocol that can be flexible with regard to specific application at individual state cancer registries where patient contact approval requirements vary.
