Testosterone therapy and cancer risks among men in the SEER-Medicare linked database
Menée à partir de données 1992-2015 des registres américains des cancers portant sur 372 696 patients atteints d'un cancer (prostate, poumon, côlon-rectum, vessie, lymphome non hodgkinien, mélanome ou oesophage) et de la base Medicare, cette étude analyse l'association entre une thérapie de remplacement de la testostérone et le risque de développer la maladie
Background: We examined associations between two forms of testosterone therapy (TT) and risks of seven cancers among men. Methods: SEER-Medicare combines cancer registry data from the Surveillance, Epidemiology, and End Results programme with Medicare claims. Our population-based case–control study included incident cancer cases diagnosed between 1992–2015: prostate (n = 130,713), lung (n = 105,466), colorectal (n = 56,433), bladder (n = 38,873), non-Hodgkin lymphoma (n = 17,854), melanoma (n = 14,241), and oesophageal (n = 9116). We selected 100,000 controls from a 5% random sample of Medicare beneficiaries and used logistic regression to estimate odds ratios (OR) and 95% confidence intervals (CI). Results: TT was associated with lower risk of distant-stage prostate cancer (injection/implantation OR = 0.72, 95% CI: 0.60–0.86; topical OR = 0.50, 95% CI: 0.24–1.03). We also observed inverse associations for distant-stage colorectal cancer (injection/implantation OR = 0.75, 95% CI: 0.62–0.90; topical OR = 0.11, 95% CI: 0.05–0.24). Risks of distant-stage colorectal and prostate cancers decreased with time after initiating TT by injection/implantation. By contrast, TT was positively associated with distant-stage melanoma (injection/implantation OR = 1.70, 95% CI: 1.37–2.11). TT was not associated with bladder cancer, oesophageal cancer, lung cancer or non-Hodgkin lymphoma. Conclusion: TT was inversely associated with distant-stage prostate and colorectal cancers but was positively associated with distant-stage melanoma. These observations may suggest an aetiologic role for TT or the presence of residual confounding.