The association of androgen deprivation therapy with time to dementia: A large competing risk analysis.

136 Background: Androgen deprivation therapy (ADT), used for advanced prostate cancer (PC), has shown mixed associations with dementia in previous studies. We evaluated whether ADT for PC raises the risk of dementia in a large, real-world sample, hypothesizing that patients receiving ADT for PC would have higher odds of dementia compared with PC patients not treated with ADT. Methods: A de-identified open-claims real-world database (NorstellaLinq) included patients age ≥40 years, diagnosed with PC between August 1, 2015, through December 31, 2021, with follow-up data through March 31, 2023. We used target trial emulation, treating the date of PC diagnosis as day 0 and examining whether receiving ADT within 180 days was associated with risk of dementia at ≥365 days. Death was inferred for patients with no claims for ≥12 months. We used competing risk analyses, accounting for death as a competing risk. Covariates included age, presence of metastatic PC, other non-PC cancers, medical comorbidities, and race, derived using an algorithm based on Bayesian Improved Surname Geocoding. Results: Of the 1,495,181 patients who met the study criteria, 9.4% received ADT within 180 days after PC diagnosis; 16% died during the study. Dementia diagnoses after day 365 were observed among 4.2% of ADT recipients and 4.2% of patients not treated with ADT. Using ADT within 180 days after PC diagnosis was not associated with dementia (aHR adjusted hazard ratio=0.993, 95% CI 0.97–1.02, p=0.639); ADT was associated with higher adjusted risk of death (aHR=1.268, 95% CI 1.25–1.28, p=3.5e –285 ). Sensitivity analyses found similar results. Compared with White patients, Black (aHR=1.40, 95% CI 1.34–1.46, p=8.9e −54 ) and Hispanic patients (aHR=1.26, 95% CI=1.21–1.32, p=6.7e −23 ) had higher risk of dementia. In analyses examining dementia risk among patients receiving various classes of ADT, patients receiving gonadotropin-releasing hormone agonists had higher risk of dementia (aHR=1.05, 95% CI 1.02–1.090) compared to patients not receiving ADT, and patients who received androgen receptor inhibitors (aHR=0.91, 95% CI 0.86–0.95) or androgen synthesis inhibitors (aHR=0.85, 95% CI 0.77–0.95) had lower risk of dementia than patients not receiving ADT. Conclusions: Contrary to our hypotheses, we found no increased risk of dementia among PC patients treated with ADT compared with those not treated with ADT in this large, real-world sample. However, future studies should further examine the findings of higher dementia risk among Black and Hispanic patients and among patients receiving gonadotropin-releasing hormone agonists.