Serum Lipids prior to Starting Androgen Deprivation Therapy and Risk of Castration Resistant Prostate Cancer and Metastasis: Results from the SEARCH Database.

ABSTRACT

PURPOSE: We tested the association of serum lipid levels prior to androgen deprivation therapy with the risk of castration resistant prostate cancer and metastasis. MATERIALS AND METHODS: We identified 302 men in the SEARCH (Shared Equal Access Regional Cancer Hospital) database who received androgen deprivation therapy after radical prostatectomy for nonmetastatic disease, had never received statins before androgen deprivation therapy and had available serum lipid data within 2 years prior to androgen deprivation therapy. Cox proportional hazards models were used to test associations between total cholesterol (less than 200 vs 200 mg/dl or greater), low density lipoprotein (less than 130 vs 130 mg/dl or greater), high density lipoprotein (40 or greater vs less than 40 mg/dl) and triglycerides (less than 150 vs 150 mg/dl or greater) and the risk of castration resistant prostate cancer and metastasis after androgen deprivation therapy while adjusting for potential confounders. Subanalyses were restricted to men who remained statin nonusers after androgen deprivation therapy. RESULTS: Median followup after androgen deprivation therapy was 67 months. Castration resistant prostate cancer and metastasis developed in 42 and 44 men, respectively. Men with elevated cholesterol received androgen deprivation therapy in an earlier year and had longer followup and a higher rate of statin use after androgen deprivation therapy. On multivariable analysis total cholesterol and low density lipoprotein were unrelated to castration resistant prostate cancer. Low high density lipoprotein (less than 40 vs 40 mg/dl or greater) was suggestively linked to an increased risk of castration resistant prostate cancer (HR 1.86, 95% CI 0.99-3.48). The association was stronger in men who remained statin nonusers after androgen deprivation therapy (HR 3.64, 95% CI 1.45-9.17). Results for metastasis were similar to those for castration resistant prostate cancer. CONCLUSIONS: Among men with nonmetastatic prostate cancer who started androgen deprivation therapy serum cholesterol was unrelated to castration resistant prostate cancer or metastasis. Low high density lipoprotein was suggestively associated with risks of increased castration resistant prostate cancer and metastasis, particularly in statin never users. Further studies are needed to explore a potential role for lipids in prostate cancer progression after androgen deprivation therapy.