777P - High-Definition Single Cell Analysis (Hd-Sca) of Prostate Cancer (Pca) Cells in Matched Bone Marrow and Blood from Patients

Abstract

ABSTRACT Aim: The approval of multiple therapies with diverse mechanisms of action for advanced PCa highlights a need to refine methods to select patients, monitor therapy effects and develop more efficacious sequences or combinations. The bone marrow is typically enriched for tumor cells in PCa. We applied this knowledge to test the hypothesis that analysis of tumor cells in bone marrow aspirates (BMA) will complement and overcome some of the challenges associated to circulating tumor cell (CTC) evaluation (serial identification, recognition of driver among heterogeneous tumor cell populations). To establish robust single PCa cell characterization for both phenotype and genotype, we developed a novel HD-SCA assay that is applied to blood and BMA specimens. Methods: Matched peripheral blood and BMA (obtained through the posterior iliac crest) were collected synchronously from 9 individual PCa patients. Nucleated cells were stained for cytokeratins (epithelial cell marker), CD45 (leukocyte marker), and androgen receptor (AR), allowing for detailed characterization of morphology, AR expression and its subcellular distribution. For single cell sequencing, candidate tumor cells were relocated and isolated. Using whole genome amplification, high-resolution whole genome copy number variation profiles were obtained from individual cells and tumor microemboli. Results: Eight patients had progressive bone metastatic castration resistant disease and 1 patient had been newly diagnosed with high-risk non-metastatic PCa (therapy naive). The number of cancer cells per mL in the marrow (median 536, range 2-4381) exceeded that in the blood (10, 1-30) in all but the non-metastatic patient, who had 13 CTC/mL and 2 disseminated tumor cells per mL. PCa cells in BMAs showed heterogeneous expression of AR, which we found positive in 2 cases with AR-negative CTC. In the 2 cases with single cell genomic analysis available, we identified cancer sub-clones with different representation in blood versus BMA for both patients. Conclusions: Our results suggest that HD-SCA provides complementary characterization of tumor cells in the fluid phase of blood and bone marrow. Morphologic and genomic profiling of single cells in simultaneously collected marrow and blood may help identify the most relevant drivers of disease progression in the individual PCa patient. Disclosure: All authors have declared no conflicts of interest.