Simple and rapid enrichment of circulating tumor cells (CTCs) for RNAseq in metastatic castrate resistant prostate cancer (mCRPC).

person Gareth Morrison University of Southern California, Los Angeles, CA info_outline Gareth Morrison, Alexander Cunha, Nita Jojo, Zarko Manojlovic, Yucheng Xu, Peggy S. Robinson, Tanya B. Dorff, David I. Quinn, Amir Goldkorn

Authors person Gareth Morrison University of Southern California, Los Angeles, CA info_outline Gareth Morrison, Alexander Cunha, Nita Jojo, Zarko Manojlovic, Yucheng Xu, Peggy S. Robinson, Tanya B. Dorff, David I. Quinn, Amir Goldkorn Organizations University of Southern California, Los Angeles, CA, USC Keck School of Medicine Norris Comprehensive Cancer Center, Los Angeles, CA, Angle PLC, Guildford, United Kingdom, City of Hope, Duarte, CA, USC Norris Comprehensive Cancer Center, Los Angeles, CA, Division of Medical Oncology, Department of Medicine, Keck School of Medicine and Norris Comprehensive Cancer Center, University of Southern California, Los Angeles, CA Abstract Disclosures Research Funding Other Background: CTCs have the potential to reflect not only genomic alterations but also cancer-relevant transcriptomic phenotypes. However, CTC gene expression has been hampered by signal-to-noise: rare CTC-derived transcripts are drowned out by abundant leukocyte-derived RNA. To date, a few specialized labs have achieved CTC RNAseq by capturing and analyzing single cells, a laborious and expensive approach not suitable for routine analysis of numerous samples. To address this need, we developed and validated a simple, rapid method for enrichment of live CTCs for RNAseq. Methods: Blood was drawn with informed consent under an IRB-approved protocol. Prostate cancer cell line spike-in samples were used to optimize live CTC enrichment by sequential leukocyte depletion (RosetteSep, Stem Cell Technologies) and size-based enrichment (Parsortix, Angle). Cancer-specific gene expression was first measured by multiplexed prostate specific qRT-PCR and subsequently by whole transcriptome amplification (WTA, SMARTer V2, Clontech) and RNAseq. Four patient samples were similarly analyzed by enrichment and RNAseq, along with spike-in positive controls and matched unenriched buffy coat negative controls. Results: Processing “from patient to RNA” took < 3 hrs. and achieved mean CTC recovery of 30% (range 28-33%) and mean leukocyte background of 100 (range 47-179), a 100,000-fold enrichment. Prostate specific genes (AR, PSA, PSMA) were consistently detected by qRT-PCR from enriched samples but not from unenriched samples. When analyzed by RNAseq, patient samples clustered with spike-in positive controls and away from matched buffy coat controls by principle component analysis and by unsupervised hierarchical clustering. Differential gene expression (enriched vs. matched buffy coat) identified prostate cancer-relevant transcripts. Conclusions: We developed a simple and efficient method for live CTC enrichment and expression profiling, applicable to large numbers of patient samples. This approach can be used serially over time to detect known cancer-specific transcripts and to discover new gene expression signatures that reflect tumor biology and inform disease management.