Abstract Background and Objective: Nadir testosterone values following initiation of androgen deprivation therapy have been shown in several studies to be prognostic for outcome, including time to castration-resistant prostate cancer (TTCRPC) and cancer specific survival (CSS). The biologic reasons for this remain unclear. Using cryopreserved serum from the PR.7 trial of intermittent vs. continuous androgen deprivation therapy(ADT), we determined to assess the role of related sex steroids as prognostic biomarkers in men on androgen deprivation therapy for recurrent cancer post-radiotherapy. Patients and Methods: Canadian patients in the PR.7 trial randomized to the continuous arm were included. Patients were excluded who did not receive ADT(n=3), had <2 years of follow-up (n=2), received exogenous estrogens or glucocorticoids(n=5), or samples were unavailable(n=36). LC-MS/MS was performed using a validated method to simultaneously analyze ten steroids: dehydroepiandrosterone (DHEA), testosterone (T), dihydrotestosterone (DHT), androst-5-ene-β,17β-diol (5-diol), androstenedione (4-dione), androsterone (AD), estrone (E1), estradiol (E2), progesterone, and androstane-3β,17β-diol (3βdiol). Descriptive statistics and correlation analyses were performed, with longitudinal changes categorized as stable, increasing, decreasing, or mixed. The prognostic value of individual steroid tertiles as well as E1:E2, E2:T, and DHT:T ratios were assessed by Kaplan-Meier analysis and Cox proportional hazards adjusted for baseline clinical variables. Outcomes assessed were TTCRPC, CSS, and overall survival (OS). Results: A total of 219 patients were included in the analysis who had a cryopreserved serum available within 2 years of randomization, with 104 patients having 2 subsequent annual samples available for measurement. Values for DHT, T, 4-dione, AD, DHEA, and 5-diol tended to be correlated among samples. Lower DHEA and AD values were significant associated with older age as was lower DHEA and DHT values with poorer performance status. Higher tertiles of E1 and E2 were associated with sooner TTCRP (log-rank, p=0.03, p=0.02, respectively). Similar trends were seen for 4-dione in predicting TTCRPC (log-rank, p=0.07). Upon adjustment, the highest tertile of E2:T had increased hazard ratios (HR) for CSS (HR=2.36, 95% CI I: 0.90-6.21; p=0.08) and TTCRPC (HR=1.70, 95% CI: 0.90-3.23; p=0.10) relative to the lowest tertile. In an analysis of the subset of patients with longitudinal values, increasing levels of AD over time were associated with poorer CSS and OS (log-rank: p=0.04 and p <0.01, respectively). On adjusted analysis, increasing AD levels had a HR for CSS of 3.43 (95% CI: 0.63-18.67; p=0.15) and a HR for OS of 4.75 (95% CI: 1.49-15.17; p <0.01). Limitations include the number of events for some groups. Conclusions: Increased levels of E1, E2, and AD during ADT correlated with adverse TTCRPC and CSS. Serum sex steroids, including both androgens and estrogens, may act as prognostic biomarkers in men receiving ADT for recurrent prostate cancer. Further investigation is warranted to support clinical use. Citation Format: Paul Toren, Azik Hoffman, Keyue Ding, France-Hélène Joncas, Frédéric Pouliot, Yves Fradet, Éric Lévesque, Chantal Guillemette, Laurence Klotz. Serum sex steroids as prognostic biomarkers in patients receiving androgen-deprivation therapy for recurrent prostate cancer post-radiotherapy: A post hoc analysis of the PR.7 trial [abstract]. In: Proceedings of the AACR Special Conference: Prostate Cancer: Advances in Basic, Translational, and Clinical Research; 2017 Dec 2-5; Orlando, Florida. Philadelphia (PA): AACR; Cancer Res 2018;78(16 Suppl):Abstract nr A072.