Drug safety analysis - no sex differences in denosumab

Abstract

Introduction: Advocacy groups claim that women are underrepresented in medical research, resulting in unsafe treatments. Osteoporosis is a chronic disease that causes bone fragility and risk of fractures and occurs predominantly in women. Denosumab is investigated for osteoporosis treatment separately in men and women, resulting in much sex-specific data. Nevertheless, it remains unclear whether the treatment is equally safe for both sexes. The objective of this research is to evaluate potential sex differences in the safety profile of denosumab. In a broader scope, we aim to formulate a methodology for sex-specific safety analysis, transcending denosumab and becoming applicable to other medicines and subpopulations. Methods: We compared 4 groups (men and women, each with denosumab and control (placebo) treatments). There is no golden standard for the assessment of differences in safety of a medicine between the sexes, because gender differences are known to occur in control treatment, hampering the interpretation of denosumab data. Adverse event (AE) data from representative studies on denosumab were retrieved from public sources and the Dutch regulatory authority’s database. Visual summaries were made for exploration of adverse event profiles, followed by statistical analysis by Fisher’s Exact test, meta-analysis and multivariate logistic regression analysis. Results: Publicly available data lacked detail for these analyses. The regulatory authority database provided workable data from 8558 females and 1927 males for analysis. Analyses detected similar safety signals that concern sex differences in the incidence of serious adverse events under denosumab and control treatment (example: Musculoskeletal disorders, relative risk (RR) of men vs. women in control arm = 0.48 (Fisher’s Exact p-value = 0.03), RR in denosumab arm =0.43 (p = 0.013). In none of the system organ classes was serious adverse event incidence affected by both sex and treatment (example: Musculoskeletal, logistic regression: sex [Male] x treatment[denosumab] p =0.824). We constructed a new plot: the2 factor efficacy and safety scatter (2FES2), enabling identification of sex differences, treatment differences and those sex differences that are attributable to denosumab. Conclusion: Statistical analysis of representative studies does not reveal important sex differences that are attributable to the presence of denosumab. Publicly available safety data from e.g., literature, and national and European assessment reports are insufficient for sex-specific safety assessment. The newly constructed 2FES2 uses data visualization, substantiated by logistic regression. This aids in the exploration, comparison and comprehensible reporting of safety profiles, detecting sex-specific safety signals. This method could be applied to assess other safety parameters such as discontinuations, or other medicines.