Introduction CTCs are one form of liquid biopsy for measuring the pharmacodynamic effects of targeted anticancer drugs in clinical trials. In this poster, we present evidence that there are diverse phenotypes of CTCs that can be classified on the basis of established tumor and tissue markers, such as epithelial of tumor origin, sarcomas sourced, or epithelial-mesenchymal transitional CTCs, which include epithelial, endothelial and mesenchymal CTCs. The current 4-color CellSearch method is dependent on expression of EpCAM, resulting in failure of CTC isolation from ~60% of patients with advanced, disseminated cancers, and cannot be applied to the collection of CTCs generated by sarcomas or lymphomas. The portion of the evaluable population is approximately 30% for all trials, and the CTC biomarker statistical evaluation is limited by the total number of CTCs collected from each tube of blood. Another limitation of the CellSearch instrument is that the number of imaging channels available precludes simultaneous positive identification of cancer cells using validated tumor markers and measurement of pharmacodynamic biomarkers. To overcome these limitations, we have been applying the ApoStream platform (DEP-FFF separation) with fluorescence imaging and using the knowledge gained from that testing, we have recently added the 5-channel CellSearch system (Mab-coated ferrofluid) adding a fifth filter, PerCP, for specimen analysis. A “home brew” CTC kit was developed that captures circulating cells that are either EpCAM or CD146 marker positive, and it was capable of identifying high numbers of EMT+ tumor marker+ double positive cells in patient specimens. Use of both CD146 and EpCAM capture per specimen, with the addition of vimentin and the tumor markers, significantly increased the number of captured CTCs that could be classified with the 5-channel CellSearch. This approach solves a major limitation of using CTCs to monitor pharmacodynamic response by enumerating statistically significant cell numbers from patients with solid tumors of diverse histology's, using a smaller blood specimen. Characterization and Enumeration of Multiple Circulating Tumor Cell Phenotypes Using Two Distinct Platforms Establishes Presence of Epithelial-Mesenchymal Transition CTCs in Patients Lihua Wang 1 , Sonny Khin 1 , Francis Owusu 1 , Ralph E. Parchment 1 , Alice Chen 2 , Shivaani Kummar 2 , James H. Doroshow 2 , and Robert J. Kinders 1 1 Laboratory of Human Toxicology and Pharmacology, Applied and Developmental Research Directorate, Leidos Biomedical Research, Inc., Frederick National Laboratory for Cancer Research, Frederick, Maryland 21702; and 2 Division of Cancer Treatment and Diagnosis, National Cancer Institute, Bethesda, Maryland 20892 Summary and Conclusions Results References Acknowledgments 1. Lihua Wang, Thomas D. Pfister, Ralph E. Parchment, Shivaani Kummar, Larry Rubinstein, Yvonne A. Evrard, Martin E. Gutierrez, Anthony J. Murgo, Joseph E. Tomaszewski, James H. Doroshow, and Robert J. Kinders. Monitoring Drug- Induced γH2AX as a Pharmacodynamic marker in individual circulating tumor cells. Clinical Cancer Research. 2010, 16:1073-1084. 2. Priya Balasubramanian, Lihua Wang, Scott M. Lawrence, Tony Navas, Shivaani Kummar, Melinda Hollingshead, Francis Owusu, Ralph E. Parchment, Joseph E. Tomaszewski, James H. Doroshow, Robert J. Kinders. Isolation and characterization of circulating tumor cells (CTCs) from peripheral blood specimens of patients with advanced solid tumor malignancies (using ApoStream™ instrumentation). 2014 AACR Annual Meeting. Abstract and poster. April, San Diego, CA. Methods • Two platforms, ApoStream and the 5-color CellSearch system, have been applied for monitoring drug responses to new anticancer agents in early stage clinical trials by measuring drug effects in circulating tumor cells (CTCs). • A “home brew” CTC kit was developed that captures circulating cells that are either EpCAM or CD146 marker positive, and it was capable of identifying high numbers of EMT+ tumor marker+ double- positive cells in patient specimens. • There was statistical equivalency of events flagged as CTCs (CD45-/EpCAM+/CK+) in patient specimens analyzed using ApoStream or the 5-color CellSearch versus the 4-color CellSearch, and 100% concordance between the CellSearch platforms. • Tumor marker specificity was established by analyzing normal donor blood, which varied from 0 to 2 cells per ml of blood depending on the marker set employed (only CD45-negative cells were assessed). • Both platforms demonstrated the presence of high numbers of CTCs with the epithelial-mesenchymal transition (EMT) phenotype, and widely used tumor markers such as MUC1, CEA, TLE1, or proteins generated by recombination events such as ASPL-TFE3. • The presence of large numbers of CTCs from carcinoma patients expressing the EMT phenotype may be related to the intensive prior treatment (>3 prior therapies, median) of our phase 1 population. This project has been funded in whole or in part with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E. This research was supported in part by the Intramural Research Program of the NIH and the Developmental Therapeutics Program in the Division of Cancer Treatment and Diagnosis of the National Cancer Institute. The CTC specimens for this study were provided by DCTD/NCI and other clinical centers at City of Hope Medical center, USC Norris Comprehensive Cancer Center, UC-Davis Comprehensive Cancer Center, Dana Farber Cancer Institute, and Yale Cancer Center. The CellSearch 5-Channel Platform was provided by Janssen Diagnostics, and PADIS/FNLCR serves as a beta testing site for this device. CTC and Biomarker-positive CTC Enumeration Comparison Setup of CellSearch CTC Assays with Home Brew CTC Kit Used for 5-Color System Versus the 4-Color System CD45 + - excluded (leukocytes) CD45 - CK + CK - MUC1/CEA + MUC1/CEA - MUC1/CEA + MUC1/CEA - CTCs CTCs CTCs CTCs Canonical CTC Patient whole blood specimens EpCAM + CD146 captured PD marker + PD marker - PD marker + PD marker - PD marker + PD marker - PD marker + PD marker - Mesenchymal CTC unknown unknown Epithelial CTC CTCs CTCs (Veridex defined CTC) CTC Characterization and PD Marker Assessment Using the 5-Color CellSearch System CTC Characterization and PD Marker Assessment Using the 5-Color CellSearch System Patient whole blood specimens EpCAM + CD146 captured CD45 + - excluded (leukocytes) CD45 - VIM + VIM - MUC1/CEA + MUC1/CEA - MUC1/CEA + MUC1/CEA - unknown CTC CTC PD marker + PD marker - PD marker + PD marker - PD marker + PD marker - PD marker + PD marker - CTC CTC unknown unknown Mesenchymal CTC Epithelial CTC 5-Color (Home brew kit) CXC-CK kit (7.5 mL) CXC-VIMENTIN kit (2 mL) Sample ID Rare Cells (CK+, CEA/MUC1-, CD45-) Epithelial CTCs (CK+, CEA/MUC1+, CD45-) Rare Cells (VIM+, CEA/MUC1-, CD45-) Mesenchymal CTCs (VIM+, CEA/MUC1+, CD45-) RDP0763 0 0 3 1 RDP0027 1 0 0 0 RDP0044 0 0 1 0 CTC Count in Healthy Donor Blood Measured by the 5-Color System with Two Types of Modified CXC Kits 5-Color (Home brew kit) CEC-CK kit (7.5 mL) CEC-VIMENTIN kit (2 mL) Sample ID Rare Cells (CK+, CEA/MUC1-, CD45-) Epithelial CTCs (CK+, CEA/MUC1+, CD45-) Rare Cells (VIM+, CEA/MUC1-, CD45-) Mesenchymal CTCs (VIM+, CEA/MUC1+, CD45-) RDP0330 3 0 0 0 RDP0459 1 0 2 0 RDP0299 0 0 0 0 CTC Count in Healthy Donor Blood Measured by the 5-Color System with Two Types of Modified CEC Kits Triple-Negative Breast Cancer Cell Line MDA-MB-231 and Endothelial Cell Line HUVEC Were Spiked into Healthy Donor Blood and Processed by the CEC Kit Using the CellSearch System CK -AF488 DAPI CD45 -APC γH2AX -PerCP CK DAPI 1 2 3 4 5 6 7 8 9 CEA/MUC1 -PE CTC Images from Post-treatment Specimen of a Patient by the 5-Color CellSearch System CTC Images from Pre-treatment Specimen of a Patient Detected by the 5-Color CellSearch System Using the Modified CXC Kit CK -AF488 DAPI CD45 -APC CEA/MUC1 -PE VIMENTIN -PerCP CK DAPI 1 2 3 4 5 6 7 8 9 CTC Images from Post-treatment Specimen of a Patient by the 5-Color CellSearch System VIM -AF488 DAPI CD45 -APC γH2AX -PerCP VIM DAPI 1 2 3 4 5 6 7 8 9 10 CEA/MUC1 -PE Leiomyosarcoma Cell Line SK-LMS-1 and Endothelial Cell Line HUVEC Were Spiked into Healthy Donor Blood and Analyzed Using the CEC Kit by the 5-Color System Various Types of CTCs Detected by the 5-Color System Using Our Home Brew CTC Kits DAPI CD45 -APC S100 -PerCP CD105+ VIMENTIN DAPI CSC CSC CEC CSC CEC CSC CEC CEC CSC CD105-PE + VIMENTIN-PE 1 2 3 4 5 6 7 8 leukocytes γH2AX -AF488 Comparison of CTC Count Measured by Different Systems Healthy Controls (n=13) Patients (n=21) 1 10 100 1000 CD45-(CK/EpCAM/ β -cat)+ cells per mL 0 (n=15) Healthy Controls (n=13) Patients (n=21) 1 10 100 1000 CD45-(MUC1/CEA)+ cells per mL 0 (n=17) EpCAM+/CK+ phenotype isolated from carcinoma patients (7.5 mL tube) MUC1/CEA+ CTC phenotype isolated from carcinoma patients (7.5 mL tube) Putative CTC phenotype isolated from sarcoma patients (7.5 ml tube) A B C CK/EpCAM/β-cat Tumor marker: (MUC1/CEA1) + + - - + + Mesenchymal CTC marker: (Vimentin+) Epithelial CTC marker: (CK/β-cat) + - + - + + Epithelial Tumor-origin (mesenchymal?) Patients with sarcoma (n=15) Mesenchymal Epithelial Patients with carcinoma (n=21) CD45-/Tumor Marker+ Circulating Cells at Baseline, Developmental Therapeutics Clinic • ApoStream is able to isolate CTCs from patients with a range of carcinomas and sarcomas • Increased numbers of (tumor-marker positive) CTCs from carcinomas as compared to EpCAM+ capture • Heterogeneity in CTCs isolated in an EpCAM-independent manner • Individual patients show multiple phenotypes • EMT in carcinomas (CEA/MUC1+); MET in sarcomas (TLE/ASPS-1/MUC1+) • Sarcoma phenotyping is complex, requiring a large variety of tumor markers ApoStream Workflow • The application of CTCs in clinical studies is currently limited by: o Greater than 50% of patient CTCs not being EpCAM positive o Inability to move back and forth between bridge species because of low cell numbers and antibody cross-reactivity requirements o Difficulties in using isolated CTCs for downstream analysis (requires viable cells for expansion and increased purity) o Inability to isolate important EpCAM negative subpopulations, specifically EMT CTCs o Requirement for high blood volumes o Relatively low cell number collection making statistical analysis challenging • A new technology (ApoStream) addresses these limitations and provides evidence of broad phenotypic plasticity among CTCs from individual carcinoma and sarcoma patients • New generation CellSearch 5-color system confirms ApoStream results and allows for the use of fixed cells Translational Relevance Tumor cells and endothelial cells were treated by topotecan at 1 µM for 2 hours Tumor cells and endothelial cells were treated by topotecan at 1 µM for 2 hours Prostate cancer, C1D8 post-treatment, 146 CTCs /1.8 L blood Uterine cancer, C1D10, post-treatment, 123 CTCs / 2 mL blood Triple-negative breast cancer, C1D1 pre-treatment, 151 CTCs / 1.7 mL blood Patients and Sample Collection • All enrolled patients and healthy subjects gave informed consent for study inclusion and were enrolled using institutional review board–approved protocols. • Blood was drawn from metastatic prostate, breast, or other cancer patients at DCTD/NCI and other clinical centers (USC, UCD, COH, DFCI). • Blood (~10 mL) was collected into CellSave® tubes (Veridex) and processed within 96 hours. • 4-mL heparinized blood collections were used for ApoStream analysis and processed the day of collection CTC Enumeration and Identification in Patient Blood • For CellSearch, a blood sample (2 to 7.5 mL) was mixed with 6.5 to 12 mL sample buffer, and was centrifuged and processed using the CellSearch platform (Janssen) and CXC kit or CEC kit. • For ApoStream, a 4-mL blood specimen was Ficoll separated and then processed through the flow chamber, and was collected and spotted onto a Marienfeld slide for image analysis • Images captured by the 5-color system in CellTracks® Analyzer II contain objects fulfilling predetermined criteria and are automatically presented in gallery format. Final classification of cells is done independently by two operators. • Cells are classified as CTCs when morphologic features and staining patterns are consistent with that of epithelial cells (CK-PE positive, DAPI positive, CD45-APC negative, and tumor marker positive) or mesenchymal cells (Vimentin positive, DAPI positive, CD45-APC negative, and tumor marker positive). • CTCs must have a minimum size of at least 4 μm, but present with a large heterogeneity in both CTC size and morphology. • γH2AX-positive CTCs present with nuclear staining in the PerCP channel. CD105 -PE DAPI CD45 -APC γH2AX -PerCP CEA/MUC1 -AF488 CD105 DAPI leukocytes CEC CTC CTC CTC CTC CTC CTC CEC 1 2 3 4 5 6 7 8 9 EpCAM + (4-Color) 5-Color system EpCAM/CD146 n/a + + + + CK 8, 18, 19 + - + + - CEA/MUC1 n/a + + + + Vimentin n/a - - + + DAPI + + + + + CD45 - - - - - Specimen ID (CTEP #8484) Diagnosis 4-color (CTC kit) (7.5 mL blood) 5-color (CXC kit) Epithelial CTC (CK+, DAPI+, CEA/MUC1+, VIM-, CD45-) Mesenchymal CTC (CK-, DAPI+, CEA/MUC1+, VIM+, CD45-) DFCI 063 C1D1 Uterine Ca 7 55/7.5 mL 11 (1.5mL blood) 100/7.5 mL 20 (1.5 mL blood) DFCI 063 C1D8 Uterine Ca 14 105/7.5 mL 28 (2.0 mL blood) 405/7.5 mL 108 (2.0 mL blood) DFCI 063 C1D9 Uterine Ca 6 60/7.5 mL 16 (2.0 mL blood) 345/7.5 mL 92 (2.0 mL blood) DFCI 063 C1D10 Uterine Ca 1 22/7.5 mL 6 (2.0 mL blood) 439/7.5 mL 117 (2.0 mL blood) DFCI 064 C1D1 Triple- negative breast Ca 2 54/7.5 mL 13 (1.8 mL blood) 367/7.5 mL 88 (1.8 mL blood) 5-Color system 4-Color system Ferrofluid capture (Enrichment) EpCAM + CD146 EpCAM Images presented in FITC channel (Positive selection & cell type identification) CK-FITC DAPI VIMENTIN-AF488 DAPI CK-PE DAPI Leukocyte maker (Negative selection) CD45-APC CD45-APC Tumor markers (CTC characterization) CEA-PE+ MUC1-PE S100/-PE PSA-PE+PSMA-PE No PD marker (PD marker identification) γH2AX -PerCP γH2AX -FITC 1 2 3 4 5 6 7 8 910 1 2 3 4 5 6 7 8 910 1 2 3 4 5 6 7 8 910 1 2 3 4 5 6 7 8 910 1 2 3 4 5 6 7 8 910 0.1 1 10 100 1000 Tumor markers Epithelial Mesenchymal Cells per 7.5 mL 0 5-Color system 4-Color system