Single CTC characterization to identify phenotypic and genomic heterogeneity as a mechanism of resistance to AR signaling directed therapies (AR Tx) in mCRPC patients.

Genitourinary Cancer
Session Type and Session Title: 
Oral Abstract Session A: Prostate Cancer
Poster Session A: Prostate Cancer
Abstract Number: 


Poster Board Number: 
Poster Session A Board #A5
J Clin Oncol 34, 2016 (suppl 2S; abstr 163)
Howard I. Scher, Ryon Graf, Jessica Louw, Adam Jendrisak, Ann Johnson, Stephanie Greene, Angel Rodriguez, Nicole A. Schreiber, Brigit McLaughlin, Lyndsey Dugan, Martin Fleisher, Jerry Lee, Yipeng Wang, Dena C. Marrinucci, Mark Landers, Ryan Vance Dittamore; Sidney Kimmel Center for Prostate and Urologic Cancers and Memorial Sloan-Kettering Cancer Center, New York, NY; Epic Sciences, Inc., San Diego, CA; Epic Sciences, San Diego, CA; Memorial Sloan Kettering Cancer Center, New York, NY

Abstract Disclosures


Background: AR directed Tx, including Abiraterone (A) and Enzalutamide (E), prolong survival in patients with mCRPC and are FDA approved. After primary Tx with either A or E, the response to E or A is decreased as is the duration of response. NGS tissue sequencing of mCRPC shows a range of disease phenotypes and genotypes which are associated with increased tumor heterogeneity. Single site biopsy, may not represent the diversity of disease. Our objective is to develop biomarkers of heterogeneity from the analysis of single cells from patients with mCRPC. Methods: 221 blood samples from 179 unique patients (pts) were collected from pts about to begin AR Tx (n = 150) or Taxanes (n = 71). Samples were analyzed with the Epic Sciences platform. Analysis included digital pathology of 20 discrete phenotypic cell features inclusive of AR, CK, size and shape measures. 9225 single CTCs were characterized, data standardized, features clustered and categorized into 15 phenotypically distinct CTC subtypes. Individual pt samples were then analyzed for the frequency and heterogeneity (Shannon Index) of CTC subtypes and monitored for clinical endpoints. A subset of CTCs (n = 350) were individually sequenced and analyzed for clonality and CNV to assess genomic heterogeneity. Results: A diversity of CTC phenotypes were observed at each line of therapy. In AR Tx pts, high heterogeneity predicts 6mo PFS (OR = 2.57, p = 0.01) and shorter OS (HR = 4.5, p < 0.001). Separately, specific CTC phenotypes also predict 6mo PFS (OR = 5, p < 0.001) and shorter OS (HR = 5.8, p < 0.001). Heterogeneity scores were not associated with taxane resistance. NGS analysis identified subclonal drivers of disease progression including multiple private and/or clustered genomic drivers. Conclusions: Single CTC characterization of phenotype and genotype supports the concept of tumor heterogeneity as a driver of disease resistance. The heterogeneity observed supports the potential value of single cell characterization to identify rare somatic sub-clonal alterations and can aid in the development of rational therapeutic trials.