Introduction:​ The purpose of this study was to retrospectively examine predictors of fracture risk when adult patients experienced a denosumab treatment lapse or discontinuation in a real-world clinic setting.

Materials and methods:​ Eligible patients were adults who had received ≥2 doses of denosumab at an academic health center in the United States. Demographics, treatment doses, reasons for missed doses, and fractures, were collected retrospectively from electronic health records, from an 8-year period (2010–2018). The number of times each patient incurred a treatment lapse, defined as ≥240 days between two doses (excluding lapse due to discontinuation, death, or transfer of care) was computed. The occurrence of denosumab discontinuation (excluding discontinuation due to death or transfer of care), whether the patient initiated alternative therapy, and the reason for each lapse and discontinuation were collected. Cox proportional hazards models assessed characteristics associated with risk of fracture and treatment discontinuation. A logistic regression model was used to determine if cumulative amount of time off medication (i.e., cumulative lapse time) was associated with a higher likelihood of incurring a fracture.

Results:​ We included 534 patients:​ 95 % White, 86 % women, 33 % tobacco users, 13 % diagnosed with diabetes, median age 69 years (Interquartile Range (IQR):​ 62–77), and median Body Mass Index (BMI) 25 kg/m2 (IQR:​ 22–28). Thirty-six percent of patients incurred 250 lapses;​ 10 % discontinued therapy. Dental problems/procedures and logistical barriers were the most common reasons for lapses and discontinuations. Nineteen percent (n = 103) incurred a total of 190 fractures;​ of these, 121 were osteoporotic, 50 were vertebral. Risk of any, osteoporotic, and vertebral fractures were associated with off-treatment status (hazard ratio [HR] = 1.7, p = 0.043;​ HR = 2.0, p = 0.026;​ and HR = 4.2, p = 0.001, respectively) and older age (HR = 1.3, p = 0.015;​ HR = 1.5, p = 0.001;​ and HR = 1.8, p = 0.005, respectively). Older age was associated with higher risk of discontinuation (HR = 1.4, p = 0.022). There was a non-significant trend of a nonlinear association between incurring a fracture and cumulative lapse time (p = 0.087).

Conclusion:​ Denosumab treatment lapses are common, and off-treatment status may be associated with a higher risk of fractures. Clinical teams should proactively identify and address adverse effects and potential logistical barriers to reduce the risk of treatment lapses.