Modern precision oncology is built on molecular information on cancer in individual patients to select optimal treatment and improve patient outcomes. While tissue biopsy samples are widely used to characterize tumors, their use to obtain real-time information is limited by constraints on repeated sampling and their incomplete representation of the complete systemic disease. Minimally invasive liquid biopsies, including analysis of circulating tumor cells and circulating tumor DNA, have been predicted to overcome these restrictions for the past two decades. They have the potential to track heterogeneity and evolution of cancer cell clones as well as for early detection of therapy resistance, recurrence and minimal residual disease after treatment with curative intention. However, the translation of liquid biopsies into routine clinical practice has been hampered by challenges in analytical standardization and demonstration of clinical utility. In this short overview, we summarize the status of this potentially disruptive application.

Prostate cancer is among the most prevalent malignancies in men and a leading cause of cancer mortality worldwide. While localized prostate cancer is often curable, progression to metastatic and castration-resistant disease either in lymph nodes or bone/bone marrow remains the major cause of death. Understanding the genomic events that drive metastasis—particularly in treatment-naïve patients—is critical to improving early detection and individualized therapy. Bulk tumor sequencing has revealed key mutational signatures but cannot resolve the cellular heterogeneity and clonal dynamics underlying metastatic spread. Single-cell genomic approaches now enable high-resolution dissection of tumor evolution, uncovering the diversity of cancer clones across disease sites. We performed whole-genome and whole-exome sequencing on single cancer cells from a treatment-naïve patient with metastatic prostate cancer, isolating cells from the primary tumor, circulating tumor cells (CTCs), disseminated tumor cells (DTCs) in bone marrow, and metastatic bone lesions. Copy number aberrations (CNAs) and single-nucleotide variants (SNVs) were characterized to define genomic heterogeneity and infer clonal relationships. Frequent monoallelic losses in tumor suppressors (PTEN, TP53, FOXO4, STAG2) and gains in oncogenes (MTOR, RAF1, HRAS) and an angiogenic growth factor (VEGFB), were observed. Metastatic cells displayed fewer genomic alterations than CTCs or DTCs. While this observation is consistent with the hypothesis that metastatic competence may be associated with relative genomic stability, normal cell contamination of the metastatic biopsy cannot be excluded, and this interpretation should be considered preliminary. Clonal evolution analysis revealed a complex branching pattern consistent with multidirectional dissemination, suggesting bidirectional seeding between the primary tumor, circulation, and metastatic sites as one possible model of spread, though alternative explanations including phylogenetic reconstruction artefacts cannot be excluded from a single-patient study. This study provides a single-cell genomic map of metastatic prostate cancer from a treatment-naïve patient, highlighting the coexistence of diverse subclones across disease sites and supporting a multidirectional model of cancer spread.These findings raise the hypothesis that metastatic progression can emerge from multiple subclones with distinct CNA and SNV profiles. Single-cell genomic profiling of untreated tumors represents a promising approach to reconstruct clonal evolution and inform precision therapies targeting early metastatic lineages, though validation in larger patient cohorts will be required. Not applicable.

Context.—:Histopathologic lymph node (LN) status has a high prognostic impact in patients with non-small cell lung cancer (NSCLC). However, there are no detailed recommendations for LN workup beyond routine hematoxylin-eosin (H&E) sections in current guidelines. Objective.—:To systematically compare the results of routine H&E, ultrastaging (US), and immunocytology (IC) LN workup for the reporting of LN involvement in NSCLC patients. Design.—:Extensive US including 10 additional step sections with serial sections for H&E and immunohistochemistry for pancytokeratin, BerEP4, and transcription termination factor 1 (TTF1) and/or p40 was performed on 122 LNs previously assessed as metastasis free during initial pathologic workup by H&E slides. One-half of each LN had also been examined by IC after disaggregation. Results.—:Twenty-four of the initially negative 122 LNs (19.7%) were positive using US and 74 of 122 (60.7%) were positive by IC, resulting in a 3.1-fold higher detection rate by IC (P = .01, χ2 test). Comparisons between initiating colony detection by US and disseminated cancer cell density (number of disseminated cancer cells per million LN cells) revealed that a disseminated cancer cell density value of about 60 reflects metastatic colony formation. Applying this value for LN staging predicted poor outcome better than histopathologic routine, at least in univariate analysis in this relatively small NSCLC cohort. Conclusions.—:Routine histopathologic workup underestimates LN spread in NSCLC patients and occasionally misses even macrometastases. US confirms and validates the significantly higher sensitivity of IC over routine histopathologic workup. In particular, IC has the potential to improve appropriate staging and prognostic stratification of patients in the future.

Circulating tumor cells (CTCs) can serve as a liquid biopsy to gain insight into treatment responses and metastatic recurrence. Due to their rarity, the analysis of CTCs is challenging and commonly based on immunomagnetic technologies using antibodies against EpCAM. This study used mass cytometry (CyTOF®) for the identification and characterization of CTCs from longitudinally monitored metastatic breast cancer (mBC) patients. Functional analysis focused on DNA damage responses, particularly the DNA repair pathway of homologous recombination (HR) validated in BC cells from the pleura. Fifty-two blood samples from 13 mBC patients were collected for the enumeration of CTCs using CellSearch® technology, isolation of CTCs together with peripheral blood mononuclear cells (PBMCs) and of plasma. Cell-free DNA (cfDNA) from plasma was analyzed by shallow genome sequencing to determine tumor fraction (TF) and HR deficiency (HRD). CTC/PBMC mixtures were phenotyped by CyTOF® using a panel of 13 antibodies including anti-γH2AX, 53BP1, and RAD51. CyTOF® identified CTCs correlating with CellSearch®- and cfDNA-based quantifications, detected DNA damage in CTCs, and the dynamics of their HR status during genotoxic therapies. Our study shows that CyTOF®-based phenotyping of CTCs from mBC patients shows promise as a method to monitor tumor progression and HR proficiency in real time for the identification of chemoresistance.

The gold standard to diagnose leptomeningeal metastasis (LM) is to identify malignant cells in the cerebrospinal fluid (CSF) or in a leptomeningeal biopsy. Using conventional cytology its sensitivity remains low with approximately 50% and 85% for the first and the second lumbar puncture, respectively. While rare cell capture technologies like the CellSearch system reach higher sensitivity than conventional cytology, they have limitations, including capturing only cells with sufficient EpCAM antigen expression, limited antibody customization, and high costs. Once diagnosed with LMs, to date most patients are still treated with untargeted therapies with limited effectiveness and significant side effects. To address this, we developed a workflow using Sievewell 370 K microwell chips. Each chip contains 370,000 hexagonal microwells (20 μm in diameter, 25 μm deep) with two 2 μm pores at the bottom. On the chip, individual cells are separated, identified with cytokeratin antibodies, and isolated as single cells via micromanipulation for molecular analysis. 15 CSF samples from 9 LM-patients were analyzed on the microwell chips and by conventional cytology in parallel. The isolated cells’ DNA was amplified by whole genome amplification and sequenced by low pass whole genome sequencing and panel sequencing on single cell level and compared to extracranial tissue biopsies. On-chip immunostaining achieved less than 5% cell loss and over 95% single cell isolation for the MCF7 breast cancer cell line. For 10 of the 15 CSF samples, concordant results were obtained by analysis on the microwell chips and conventional cytology: 1 sample was tumor cell-negative, 9 samples were tumor cell-positive. In the other 5 samples, tumor cells were detected by analysis on the microwell chip with either inconclusive (3 samples) or negative (2 samples) results obtained by cytological analysis. With the microwell chips, 5 to 10,000 (mean 1,218; median 116) CSF-CTCs were found per mL CSF. The detection of chromosomal aberrations confirmed the malignant origin of detected tumor cells, including those from samples that were tumor cell-negative in conventional cytology. A high level of clonality was observed among the tumor cells in the CSF of each patient, yet these CSF tumor cells often exhibited an aberration pattern with distinct differences compared to the corresponding extracranial metastases or primary tumors. In addition to mutations shared by corresponding tissue biopsies, panel sequencing revealed private mutations of CSF-CTCs in resistance related genes and tumor driver genes. In conclusion, the microwell-based CSF-CTC detection appears superior to conventional cytology. The high level of clonality indicates a very close evolutionary relationship within the CSF-CTCs and a monoclonal origin of LMs. Private mutations in resistance related genes and tumor driver genes could impact drug susceptibility and therapy resistance. Therefore, the CSF-based liquid biopsy should be considered for genetic profiling of LMs, potentially improving diagnosis, treatment monitoring, and targeted therapy selection. Citation Format: André Franken, Martin Schramm, Barbara Alberter, Jens Warfsmann, Bernhard Polzer, Franziska Meier-Stiegen, Natalia Krawczyk, Michael Sabel, Marion Rapp, Eugen Ruckhäberle, Tanja Fehm, Hans Neubauer. Detection, isolation, and genetic characterization of circulating tumor cells from the cerebrospinal fluid of breast cancer patients with leptomeningeal metastasis [abstract]. In: Proceedings of the San Antonio Breast Cancer Symposium 2024; 2024 Dec 10-13; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2025;31(12 Suppl):Abstract nr P4-01-27.

Abstract Introduction: Recent studies on disseminated cancer cells (DCCs) in Non-Small-Cell Lung Cancer (NSCLC) patients demonstrated their clinical relevance even in early tumor stages. However, their molecular and phylogenetic link to the primary tumor is currently unclear. Certain genetic alterations, such as TP53 or KRAS mutations, have been proposed to be clonal events and drivers of dissemination and metastasis. In this study, we analyzed the genome and transcriptome of single DCCs to identify the characteristics of early DCCs and their translational clinical value. Materials and methods: From 2011 to 2019 we prospectively collected and analyzed bone marrow aspirates and lymph node samples from 296 newly diagnosed NSCLC patients undergoing tumor resection with curative intention. Presence of DCC was examined by i) immunocytology (IC) for Cytokeratin (clone A45-B/B3, bone marrow, BM-DCCs) or EpCAM (clone BerEp4, lymph node, LN-DCCs) and ii) immunofluorescent (IF) staining for EpCAM in a live cell suspension (clone HEA-125) from BM or LN. Single cells were isolated using micromanipulation and subjected to Whole Genome Amplification or combined Whole Genome/Transcriptome Amplification for IC and IF, respectively. Sanger sequencing for known hotspot mutations in TP53 and KRAS was performed on single DCCs. Gene expression profiles of DCCs were analyzed by single cell RNA sequencing and the results were compared with clinical data with a mean follow-up of 2.93 years (range 0.03-7.22). Results: LN-DCCs were found in 44.8% and 59.6% by IC and IF, respectively. CKIC pos BM-DCCs were found in 45.2%, while EpCAMIF pos BM-DCCs were present in 52.8%. LN-DCCs correlated with reduced tumor-specific survival (TSS, p=0.009) and progression-free survival (PFS, p=0.025). EpCAMIF pos, but not CKIC pos BM-DCCs correlated with reduced TSS (p=0.003) and PFS (p=0.083). Genetic analysis of single DCCs revealed that TP53 or KRAS mutations could be found in 3.6% and 2.8% of analyzed BM-DCCs and 5.0% and 29.3% of LN-DCCs for IF and IC, respectively. Single cell RNA-seq showed distinct transcriptomic profiles for LN- and BM-DCCs, and identified a DCC phenotype with high stemness score in the bone marrow. Further analysis of potential transcriptome-based biomarkers identified a novel candidate marker for activated EpCAMpos BM-DCCs that strongly correlates with PFS and TSS (p<0.001). Conclusions: DCCs in early-stage NSCLC have high prognostic value and predict reduced PFS and TSS. While genetic alterations in the TP53 and KRAS gene were not detected in the majority of analyzed DCCs, transcriptomic phenotypes provided relevant information and identified a subset of activated DCCs that correlate strongly with relapse and poor outcome. Citation Format: Tobias Mederer, Daniel Spitzl, Felix Elsner, Bernhard Polzer, Michael Ried, Reiner Neu, Tobias Robold, Hans-Stefan Hofmann, Christoph A. Klein. Comprehensive analysis of single disseminated cancer cells identifies a stem-like phenotype with strong prognostic impact [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3778.

The analysis of circulating tumor cells (CTCs) has shown potential for detection of cancer spread, prognosis, therapeutic target selection, and monitoring of treatment response. CTCs can be obtained repeatedly by simple blood draws as so-called "liquid biopsy." Thus, they can serve as a surrogate material for primary or metastatic tissue biopsies. In addition, isolation of CTCs provides the possibility to investigate those cells which may hold the (molecular) traits responsible for metastatic progression and ultimately patient death. As such, CTCs represent a target of utmost importance in cancer research and therapy. In this chapter, we describe a workflow for the enrichment of CTCs with the FDA-cleared CellSearch® system followed by the isolation of single CTCs using the DEPArray™ technology enabling further molecular single-cell analyses.

While cell‐free liquid biopsy (cfLB) approaches provide simple and inexpensive disease monitoring, cell‐based liquid biopsy (cLB) may enable additional molecular genetic assessment of systemic disease heterogeneity and preclinical model development. We investigated 71 blood samples of 62 patients with various advanced cancer types and subjected enriched circulating tumor cells (CTCs) to organoid culture conditions. CTC‐derived tumoroid models were characterized by DNA/RNA sequencing and immunohistochemistry, as well as functional drug testing. Results were linked to molecular features of primary tumors, metastases, and CTCs; CTC enumeration was linked to disease progression. Of 52 samples with positive CTC counts (≥1) from eight different cancer types, only CTCs from two salivary gland cancer (SGC) patients formed tumoroid cultures (P = 0.0005). Longitudinal CTC enumeration of one SGC patient closely reflected disease progression during treatment and revealed metastatic relapse earlier than clinical imaging. Multiomics analysis and functional in vitro drug testing identified potential resistance mechanisms and drug vulnerabilities. We conclude that cLB might add a functional dimension (to the genetic approaches) in the personalized management of rare, difficult‐to‐treat cancers such as SGC.

Comprehensive genome-wide analyses of single cells represent an important tool for clinical applications, such as pre-implantation diagnostic and prenatal diagnosis, as well as for cancer research purpose. For the latter, studies of tumor heterogeneity, circulating tumor cells (CTCs), and disseminated cancer cells (DCCs) require the analysis of single-cell genomes. Here we describe a reliable and robust array-based comparative genomic hybridization (aCGH) protocol based on Ampli 1™ whole genome amplification that allows the detection of copy number alterations (CNAs) in single cancer cells as small as 100 kb.

Die Analyse von Einzelzellen birgt großes Potenzial für die Diagnostik und die Therapie verschiedenster Erkrankungen. In der Onkologie werden bereits im Rahmen der Liquid Biopsy einzelne zirkulierende Tumorzellen im Blut charakterisiert, um Krankheitsverläufe zu überwachen oder die Therapie optimal anzupassen. Die Weiterentwicklung der Einzelzellanalyse beispielsweise durch Spatial-Omics-Technologien sowie die bioinformatische Auswertung verschiedener Informationsebenen lässt hoffen, dass in Zukunft bei komplexen physiologischen Prozessen pathogene Abweichungen in nur wenigen Zellen frühzeitig identifiziert werden können.

The role of circulating tumor cell (CTC) clusters in the metastatic dissemination process is gaining increased attention. Besides homotypic clusters, heterotypic clusters that contain tumor cells admixed with normal cells are frequently observed in patients with solid tumors. Current methods used for cluster detection and enumeration do not allow an accurate estimation of the relative fractions of tumor cells. Here we describe a method for estimating tumor fraction of clusters including isolation and collection of single clusters, assessment of copy number alterations of single clusters by low-pass whole genome sequencing, and bioinformatic analysis of sequencing data.

Abstract Introduction: Liquid biopsy enables minimally invasive molecular profiling of systemic cancer for diagnosis, therapy selection and longitudinal monitoring of minimal residual disease. Cell-based liquid biopsy methods are particularly well suited to examine the extent of genetic heterogeneity and clonal diversity among metastatic cancer cells, which are believed to be the major determinants of therapeutic failure. However, isolation and molecular profiling of circulating tumor cells (CTCs) from body fluids is challenging due to methodological constrains. Methods: To overcome these challenges, we have developed a workflow for detection, isolation, and mutational profiling of CTCs. The workflow utilizes the FDA-approved CellSearch System® and a novel method for single-cell Integrated Mutation Profiling of Actionable Cancer Targets (scIMPACT). Our workflow was optimized for reliable detection of copy number variations (CNVs) and single nucleotide variations (SNVs) in single CTCs. The method was then applied to 31 CTCs alongside with 15 matched germline samples. Results: After implementation of a customized Bayesian neural network based algorithm, accounting for biases introduced during amplification of single-cell DNA and prioritizing the detected mutations, we achieved a sensitivity of 91% and a specificity of 96%, AUC = 0.98. The CTC analysis revealed presence of CNVs affecting known oncogenes and tumor suppressor genes (including MYC, CCND1, FGF4, TP53 and RB1) as well as SNVs classified as pathogenic (e.g., ARID1A Q775*). Notably, in the examined specimens most of the chromosomal aberrations (51-82%) were clonal across all cells of a given patient. In contrast, most of the detected somatic SNVs (67-96%) were sub-clonal and thus present only in a subset of CTCs. Conclusion: In summary, the novel scIMPACT workflow enabled reliable and accurate genomic profiling of patient-derived CTCs in late-stage breast cancer patients. Our proof-of-concept study revealed evidence of genetic diversity of SNVs among the examined CTCs, which might have been acquired late in progression but relevant for therapy escape in the studied cases. Further analyses are still needed for assessing the generalizability of our finding as well as for deeper understanding of the molecular basis orchestrating the establishment and maintenance of heterogeneous genotypes of CTCs in breast cancer. Citation Format: Zbigniew T. Czyż, Clara Chaiban, Adithi Ravikumar Varadarajan, Cäcilia Köstler, Vadim Dechand, Jonas Grote, Thomas Ragg, Bernhard Polzer, Jens Warfsmann, Christoph A. Klein. Genomic profiling with the novel single-cell IMPACT assay reveals evidence of genetic heterogeneity among CTCs in late-stage breast cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3692.

Abstract Introduction: Leptomeningeal metastases (LM) are diagnosed in up to 10% of metastatic cancer patients, the most frequent cancers being carcinomas of breast and lung or melanoma. Due to a dismal prognosis and an overall median survival of only four months new approaches for molecular analysis could help to identify individual treatment options. Material and Methods: In this study, we collected 28 cerebrospinal fluid (CSF) samples from patients with different primary tumors but suspected LM. CSF was divided in two parts for standard neuropathologic cytomorphology and detection of circulating tumor cells (CTCs) by using the CellSearch® CTC Kit, respectively. Subsequently, single Cytokeratin-positive CTCs were isolated for whole genome amplification by Ampli1TM and molecular analysis. Additionally, we were able to isolate cell-free DNA (cfDNA) from CSF in a subset of patients which we amplified by an adapted Ampli1TM library protocol. In a small subset of patients, we collected slices of formalin-fixed paraffin embedded (FFPE) primary tumors and systemic metastases for dissociation and isolation of pure subpopulations by DEPArrayTM technology. Isolated CTCs, cfDNA and subpopulations of tumor and stromal cells were analyzed for copy number variations (CNV) applying the Ampli1TM LowPass protocol on the Illumina MiSeqTM platform. Results: By comparing results from classical cytomorphology and CellSearch® workflow, we could find high concordance (24/28 samples) with a slightly higher tumor cell detection rate by CellSearch® (11 positive cases vs. 9 by cytomorphology). Importantly, we could confirm the malignant origin of isolated cells by CNV analysis in all patients positive in CellSearch® assay. Moreover, by comparing single cell data with cfDNA analysis of a subset of 20 samples, we could also detect a higher sensitivity for tumor detection in CTC-based analysis (8/20 cases positive) in comparison to cfDNA-based analysis (5/20 cases positive). In a comprehensive analysis of multiple samples of an individual patient, including CTCs and cfDNA from CSF at different time points, CTCs from blood and biopsies from multiple visceral metastasis, we could identify divergent genomic evolution between tumor cells from the central nervous syste Conclusion: In this proof-of-concept study we show that a liquid biopsy-based approach for single cell analysis from CSF holds potential for innovative diagnostic assays for patients with suspected LM. Combining CTC quantification with molecular single cell analysis could enable us to define predictive tests to select novel treatment options for patients with LM in the future. Citation Format: Caecilia Koestler, Giancarlo Feliciello, Florian Lueke, Kathrin Weidele, Saida Zoubaa, Peter Hau, Tobias Pukrop, Markus J. Riemenschneider, Christoph A. Klein, Bernhard Polzer. Single cell analysis from CSF of patients with suspected leptomeningeal metastasis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3709.

Legends for Supplementary Figures S1-S11. Supplementary Figure S1 - Workflow employed in this study. Supplementary Figure S2 - Photographs of cell lines stained with isotype antibodies. Supplementary Figure S3 - Box-plot representing BM cell numbers before and after immunomagnetic selection, number of screened cells and number of isolated EpCAM+ cells. Supplementary Figure S4 - Photographs of cell lines stained for EpCAM. Supplementary Figure S5 - Photographs and flowcytometric analysis of BM of control patient spiked with DU145 cells. Supplementary Figure S6 - Expression analysis of BM cell pools. Supplementary Figure S7 - QC of WTA and overview of epithelial transcript expression. Supplementary Figure S8 - Results of expression profiling of all samples. Supplementary Figure S9 - Analysis of WTA controls for HBA2 transcript. Supplementary Figure S10 - CGH profiles of analysed single cells. Supplementary Figure S11 - Establishing conditions for specific amplification of selected fragments by PCR and confirmation of identity of amplified fragments.

Ablative radiotherapy is a highly efficient treatment modality for patients with metastatic prostate cancer (PCa). However, a subset of patients does not respond. Currently, this subgroup with bad prognosis cannot be identified before disease progression. We hypothesize that markers indicative of radioresistance, stemness and/or bone tropism may have a prognostic potential to identify patients profiting from metastases‐directed radiotherapy. Therefore, circulating tumor cells (CTCs) were analyzed in patients with metastatic PCa (n = 24) during radiotherapy with CellSearch, multicolor flow cytometry and imaging cytometry. Analysis of copy‐number alteration indicates a polyclonal CTC population that changes after radiotherapy. CTCs were found in 8 out of 24 patients (33.3%) and were associated with a shorter time to biochemical progression after radiotherapy. Whereas the total CTC count dropped after radiotherapy, a chemokine receptor CXCR4‐expressing subpopulation representing 28.6% of the total CTC population remained stable up to 3 months. At once, we observed higher chemokine CCL2 plasma concentrations and proinflammatory monocytes. Additional functional analyses demonstrated key roles of CXCR4 and CCL2 for cellular radiosensitivity, tumorigenicity and stem‐like potential in vitro and in vivo. Moreover, a high CXCR4 and CCL2 expression was found in bone metastasis biopsies of PCa patients. In summary, panCK + CXCR4 + CTCs may have a prognostic potential in patients with metastatic PCa treated with metastasis‐directed radiotherapy.

The cellular basis of the apparent aggressiveness in lung cancer is poorly understood but likely associated with functional or molecular features of disseminated cancer cells (DCCs). DCCs from epithelial cancers are mostly detected by antibodies directed against histogenetic markers such as cytokeratin or EpCAM. It has been argued that marker‐negative metastatic founder cells might escape detection. We therefore used ex vivo sphere formation for functional detection of candidate metastasis founders. We generated cell suspensions from 199 LN samples of 131 lung cancer patients and placed them into non‐adherent cell culture. Sphere formation was associated with detection of DCCs using EpCAM immunocytology and with significantly poorer prognosis. The prognostic impact of sphere formation was strongly associated with high numbers of EpCAM‐positive DCCs and aberrant genotypes of expanded spheres. We also noted sphere formation in patients with no evidence of lymphatic spread, however such spheres showed infrequent expression of signature genes associated with spheres from EpCAM‐positive samples and displayed neither typical lung cancer mutations (KRAS, TP53, ERBB1) nor copy number variations, but might be linked to disease progression >5 years post curative surgery. We conclude that EpCAM identifies relevant disease‐driving DCCs, that such cells can be expanded for model generation and that further research is needed to clarify the functional and prognostic role of rare EpCAM‐negative sphere forming cells.

Supplementary Tables S1-S4. Supplementary Table S1 - Sequences of primers used in this study. Supplementary Table S2 - Detection rate of EpCAM+ cells and the number of isolated and analyzes EpCAM+ single cells. Supplementary Table S3 - Numbers and frequencies of samples expressing analyzed transcripts. Supplementary Table S4 - Distribution of cells expressing or co-expressing EPCAM and KRT transcripts.

Background Circulating tumour cells (CTCs) are mainly enriched based on the epithelial cell adhesion molecule (EpCAM). Although it was shown that an EpCAM low-expressing CTC fraction is not captured by such approaches, knowledge about its prognostic and predictive relevance and its relation to EpCAM-positive CTCs is lacking. Methods We developed an immunomagnetic assay to enrich CTCs from metastatic breast cancer patients EpCAM independently using antibodies against Trop-2 and CD-49f and characterised their EpCAM expression. DNA of single EpCAM high expressing and low expressing CTCs was analyzed regarding chromosomal aberrations and predictive mutations. Additionally, we compared CTC-enrichment on the CellSearch system using this antibody mix and the EpCAM based enrichment. Results Both antibodies acted synergistically in capturing CTCs. Patients with EpCAM high-expressing CTCs had a worse overall and progression-free survival. EpCAM high- and low-expressing CTCs presented similar chromosomal aberrations and mutations indicating a close evolutionary relationship. A sequential enrichment of CTCs from the EpCAM-depleted fraction yielded a population of CTCs not captured EpCAM dependently but harbouring predictive information. Conclusions Our data indicate that EpCAM low-expressing CTCs could be used as a valuable tumour surrogate material—although they may be prognostically less relevant than EpCAM high-expressing CTCs—and have particular benefit if no CTCs are detected using EpCAM-dependent technologies.

Background Circulating tumor cells (CTCs) are considered as precursors of metastatic spread and can act as prognostic and predictive biomarkers. Their enrichment is mainly based on immunomagnetic technologies relying on antibodies against EpCAM. Although it was shown that an EpCAM low-expressing fraction is not captured by this, knowledge about its clinical relevance and relation to EpCAM-positive CTCs is still lacking.