Circulating tumor cells (CTCs) N-terminal androgen receptor expression to identify patients (pts) with castrate resistant prostate cancer (CRPC) who are more sensitive to chemotherapy.

5034 Background: Loss of the retinoblastoma tumor suppressor (RB) function was identified as a major means to develop CRPC; the expression of the androgen receptor (AR) is under stringent RB control; and tumors devoid of RB function are hypersensitive to treatment with chemotherapy. Exploratory analysis evaluated baseline N-terminal AR expression in CTCs in men with chemotherapy-naïve CPRC and correlated to changes in PSA, leading us to inquire if this biomarker may identify pts sensitive to chemotherapy. Methods: In a multicenter phase II randomized trial of approved doses of abiraterone acetate/prednisone (AA-Arm 1) or combination AA and standard doses of cabazitaxel (AA/CBZ-Arm 2). Patients on AA received CBZ upon progression. Baseline CTCs were obtained on all pts and expression of N-terminal AR expression was performed by Epic Sciences. Positive AR N-terminal expression (AR) was based on the presence of at least 1 CTC or CK cell with AR N-terminal signal expression above the 3.0 positivity threshold. Serial PSAs were determined at baseline and every 3 weeks with routine labs and imaging every 12 weeks. Results: To date, 42 of 80 pts have been enrolled: 22 pts to AA, and 20 pts to AA/CBZ. Both regimens were well tolerated with 8/42 (19%) pts experiencing treatment-related grade 3 or 4 toxicities. Blood from 35 patients underwent CTC analysis. Seventy-seven percent of pts (27/35) had detectable CTCs; 11 of 35 pts (31%) had AR overexpression. Of the pts with AR+ CTCs, 1/5 pts treated with AA, and 5/6 pts treated with AA/CBZ had a PSA decline > 50% from baseline. Conclusions: Real-time CTC analysis of N-terminal AR expression was feasible and data suggests that this may identify a cohort of pts who may benefit from the combination of CBZ with AA. Further studies are ongoing to evaluate whether cellular heterogeneity and RB expression in CTCs play a role in identifying pts who would benefit from chemotherapy. The trial is coordinated by the Prostate Cancer Clinical Trials Consortium, LLC and funded by Sanofi US Services Inc. and Prostate Cancer Foundation. Clinical trial information: NCT02218606.

Duke Scholars

Author Susan Halabi Biostatistics & Bioinformatics, Division of Biostatistics