Gene Expression Of Circulating Tumor Cells Is Predictive Of Treatment Response In Patients With Advanced Prostate Cancer

Background: Genome and transcriptome-based analysis has begun to reshape the approach to prostate cancer (PC). Two different gene expression signatures have shown that PC can be divided into 3 subclasses reflecting luminal-basal biology. These subtypes point toward biological drivers that may strongly influence how care should be personalized including optimization of androgen receptor targeted therapy. The majority of work done in this area has been based on tissue-based gene expression. With the advent of newer nanotechnology platforms for isolation of circulating tumor cells (CTCs), profiling of PC gene expression from blood is now possible. Methods: We recruited 34 patients with metastatic castration resistant PC who had available blood specimens prior to initiation of androgen receptor signaling inhibitor (ARSI, e.g. abiraterone, enzalutamide and apalutamide) therapy. We combined variations of the NanoVelcro Assays (thermoresponsive, click-chemistry) allowing for capture and release of CTCs with intact mRNA. Gene sets from the PCS and PAM50 signatures were re-reviewed to optimize signal detection in the blood and enriched for genes upregulated in PC. The NanoString nCounter platform was applied to profile the resulting genes. A pilot study was conducted using banked samples available through the Urologic Oncology Program Blood and Biospecimen Bank at Cedars-Sinai Medical Center. Results: The final assay was tested in banked blood samples and provided classifications of patients that associated with clinical responsiveness to therapy. Validation was conducted to examine the performance of the CTC-specific PCS/PAM50 panel in public databases (including Prostate Cancer Transcriptome Atlas and GenomeDx). Our pilot study showed that the median overall survival was significantly worse in PCS1 patients. Conclusions: This study shows initial proof of principle that genomic classification in blood is possible using contemporary tool for blood component isolation and RNA profiling. Additional technical and clinical validations are needed prior to widespread implementation, but these methods may make it possible to increase the utilization of genomic classifiers in clinical studies and in practice. Citation Format: PAI-CHI TENG, Minhyung Kim, Yu Jen Jan, Jie-Fu Chen, Nu Yao, Gina C. Chu, Pin-Jung Chen, Jasmine J. Wang, Yi-Te Lee, Yazhen Zhu, Leland W. Chung, Felix Y. Feng, Michael R. Freeman, Sungyong You, Hsian-Rong Tseng, Edwin M. Posadas. Gene expression of circulating tumor cells is predictive of treatment response in patients with advanced prostate cancer [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5447.