Background:​ Teriparatide, a recombinant human parathyroid hormone analogue, is associated with increased bone mineral density and a decreased risk of fractures. A dose-dependent increase in the incidence of osteosarcoma was observed in toxicology studies conducted in rats. The primary objective of this study was to estimate the incidence of osteosarcoma over a 10-year period among teriparatide-treated patients versus patients unexposed to teriparatide with osteoporosis and patients in the general population using national pharmacy dispensing data linked with data from participating state cancer registries (SCRs) in the US.

Methods:​ Patients aged 18 years or older with a dispensed teriparatide prescription formed two different cohorts:​ Teriparatide-Osteoporosis (Teriparatide-OP) was formed by matching teriparatide patients to unexposed patients with osteoporosis and Teriparatide-General Population (Teriparatide-GP) was formed by matching teriparatide patients to general population patients with a dispensed prescription for a medication other than teriparatide. Matching was performed using select demographics and other variables. Study cohorts were linked to SCR data to ascertain osteosarcoma status. To account for missing outcome data from non-participating SCRs, two analytic approaches were used:​ the first adjusted the person-time at-risk using a coverage fraction and the second restricted the analyses to patients from states with participating SCRs.

Results:​ There were 18 osteosarcoma cases across four study cohorts:​ the same three cases in the Teriparatide-OP and Teriparatide-GP cohorts, six cases in the Osteoporosis cohort, and nine cases in the General Population cohort. For the analysis using the coverage fraction the incidence rate ratio (IRR) comparing the Teriparatide-OP and Teriparatide-GP cohorts to the Osteoporosis and General Population cohorts was 1.0 (95% CI:​ 0.2, 4.5) and 1.3 (95% CI:​ 0.2, 5.1), respectively. When restricting the analysis to patients from states with participating SCRs, the IRR was 0.6 (95% CI:​ 0.1, 3.6) and 0.8 (95% CI 0.1, 4.0), respectively.

Conclusion:​ The estimates of association between teriparatide and osteosarcoma were imprecise due to the small number of observed osteosarcoma cases. However, the incidence of osteosarcoma observed in each study cohort was within the expected range given background rates for the US general population. The evidence generated by this study, in conjunction with other real-world studies evaluating the risk of osteosarcoma, was used to support changes to the US teriparatide label (including removal of the black box warning regarding potential risk of osteosarcoma) and expand treatment options for patients with osteoporosis.