Incidence of osteonecrosis of the jaw in patients with bone metastases treated sequentially with bisphosphonates and denosumab.

OBJECTIVES: Osteonecrosis of the jaw (ONJ) is a serious adverse event of bone resorption inhibitors (BRIs), such as bisphosphonates and denosumab. Bisphosphonates and denosumab inhibit osteoclast function through different pharmacological effects and bisphosphonates are retained in bone for several months to years. Sequential treatment with bisphosphonates and denosumab might lead to an overlapping treatment effect, due to the addition of the effect of denosumab on the residual bisphosphonate effect. Therefore, the aim of our study was to investigate if switching from denosumab to bisphosphonates is associated with a higher incidence of ONJ.
METHODS: We retrospectively reviewed records of patients with solid tumors and bone metastases treated with denosumab after prior treatment with bisphosphonates at the University Hospitals Leuven (sequential group). Patients treated with denosumab or bisphosphonates alone were used as control groups.
RESULTS: We identified 110 patients sequentially treated with bisphosphonates and denosumab with a median total BRI exposure of 36 months (sequential group). Median bisphosphonates exposure was 16 months and median denosumab exposure was 13 months. About 299 patients were included in the bisphosphonates control group with a median bisphosphonate exposure 19 months. About 6.7% (20/299) of patients developed ONJ. About 240 patients were included in the denosumab control group with a median denosumab exposure 17.5 months. About 10.0% of patients (24/240) developed ONJ. In the sequential group, 15.5% of patients (17/110) developed ONJ. The incidence of ONJ was 1.8% (2/110), 6.3% (6/99), 4.9% (4/82), 5.6% (3/54), and 3.4% (1/29), respectively in the first, second, third, fourth, and fifth year of BRI exposure, an ONJ-incidence similar to ONJ-incidence in the denosumab control group. In a time-to-ONJ-analysis, the curves of the sequential group and the denosumab control group were overlapping. In the sequential group, most of the ONJs occurred in the first year of denosumab exposure and in a matched control group analysis, with correction for median BRI-exposure, ONJ cases tend to occur earlier in the sequential group compared to ONJ cases in the bisphosphonates group.
CONCLUSION: Cancer patients with bone metastases treated with BRIs seem to have a slightly higher risk of ONJ early after switching from bisphosphonates to denosumab compared to patients remaining on bisphosphonates. Nevertheless, based on the global ONJ-incidence, the switch from bisphosphonates to denosumab can be considered as safe as an equivalent exposure to denosumab from the start on.