The emerging role of circulating tumor cells in breast cancer


  • Savitri Krishnamurthy MD

    Corresponding author
    1. Department of Pathology, The University of Texas MD Anderson Cancer Center, Houston, Texas
    • Department of Pathology, Unit 0053, The University of Texas MD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX 77030

    Search for more papers by this author
    • Fax: (713) 792-3738


The evaluation of circulating tumor cells in peripheral blood of patients with breast cancer can be performed using different platforms. Standardization and comparison of the platforms to each other and their use in prospective clinical trials is needed for incorporation of circulating tumor cell assays into clinical practice.

The first documentation of circulating tumor cells (CTCs) dates back to 1869, when Ashworth described a case in which tumor cells were found in the blood of a patient after death.1 Since then, there have been several reports of CTCs in the blood of patients with various solid tumors, designated by different terms such as carcinocythemia and cancer cell leukemia.2 Although CTCs are known to circulate in the peripheral blood of patients with breast, prostate, colorectal, and lung carcinomas, they are rare in healthy subjects or in patients with nonmalignant disease.3 Early reports recognized the presence of CTCs as an indicator of poor clinical outcome in patients with epithelial malignancies.

The relative number of CTCs in the peripheral blood of patients with solid tumor is generally far lower than the relative number of other types of cells, with CTCs making up only about 1 in 106 to 1 in 107 of all cells in the blood. The rarity of CTCs in blood makes their detection extremely challenging. Until recently, the lack of robust and reproducible methods of detection for CTCs has been the main hindrance to exploring the clinical utility of CTCs in the management of patients with solid tumors.

The natural history of CTCs is not entirely clear. These cells are most likely part of a central event in the complicated process of metastasis, in which the CTCs separate from the main tumor, enter the bloodstream, and eventually migrate to distant organs to develop secondary tumors.4 There is sufficient evidence to indicate that persistence of these cells in the blood is most likely linked to metastatic relapse. However, a proportion of these cells may be apoptotic and are eventually removed from circulation, and some of these cells may remain in a dormant state for indefinite periods without ever resulting in metastasis.5, 6 The factors contributing to or delaying metastatic relapse, however, are not entirely clear. Therefore, although elevated CTC levels appear to be associated with tumor progression, CTCs by themselves do not predict metastasis, and many patients with elevated CTC levels may be otherwise doing well.

Technological advances in recent years have made it possible to detect and enumerate CTCs in peripheral blood; several different technologies are currently available for this purpose. Because of the rarity of CTCs in the blood, the majority of these technologies use some kind of enrichment step followed by a detection step. Methods such as density gradient separation, differential centrifugation, and immunomagnetic separation have been used for enrichment. The most commonly used of these is immunomagnetic separation with antibody-based magnetic capture, in which magnetically bound antibodies are directed against the epithelial cell adhesion molecule (EpCAM), so that EpCAM-positive cells are selected or against clusters of differentiation 45 (CD45) so that CD45-positive cells are not selected. Various techniques are used for the detection of CTCs after immunomagnetic capture, including direct antibody-based methods in which antibodies are directed against cytokeratin (CK), such as immunocytochemistry, immunofluorescence, or flow cytometry, and indirect nucleic acid–based methods that measure messenger RNA (mRNA) transcript expression levels by reverse transcriptase polymerase chain reaction (RT-PCR).

CellSearch (Veridex LLC, Raritan, NJ), a semiautomated, fluorescence-based microscopic assay, is the only standardized, objective test approved by the US Food and Drug Administration for the detection of CTCs in patients who have metastatic breast, prostate, and colorectal cancer. This assay is a combined enrichment and detection system that uses EpCAM antibodies attached to microscopic iron particles (ferrofluids) for enrichment.7 Once the cells that express EpCAM bind to the antibodies, magnets pull them from the blood and separate them from other nucleated cells. The enriched cells are subsequently subjected to immunofluorescent staining using CD45 and CK antibodies (CK8, CK18, CK19) for detection. The cells are recognized as CTCs if they are CK-positive and CD45-negative and have the signatures of malignant cells, including large size and large nuclei with or without nucleoli. Cells are analyzed using an automated microscope and counted using a cell spotter analyzer (Veridex). This fluorescence-based microscopic assay has been shown to produce highly reproducible results, with good interobserver and intraobserver concordance.

Another CTC detection test is the AdnaTest Breast Cancer Select/Detect (AdnaGen, Langenhagen, Germany), which is based on the analysis of tumor-associated mRNA isolated from CTCs after immunomagnetic separation that uses EpCAM and mucin 1 (MUC1) as the targets.8 For detection of CTCs, multiplex RT-PCR is performed to analyze the relative expression levels for 3 tumor-associated transcripts: human epidermal growth factor 2 (HER2), MUC1, and GA733-2. This test is scored positive if at least 1 of the PCR transcript products for the 3 markers is detected at a concentration of >0.15 ng/μL. Other CTC detection methods include the membrane microfilter assay, epithelial immunospot assay, fiber-optic array scanning, laser scanning cytometry, and automated immunomagnetic separation (MagSweeper).9

There are distinct advantages and disadvantages to morphologic versus molecular methods. Technologies based on morphologic evaluation are very specific and allow interphase fluorescence in situ hybridization (FISH) to be performed, but they are generally less sensitive.8, 10 Molecular methods, which use a wide range of epithelial-specific and breast cancer–specific markers such as CK19, MUC1, and mammaglobin, are generally more sensitive and allow for the evaluation of multiple markers simultaneously, but they may generate false positive results due to low expression of the tested marker in noncancerous cells.8, 10, 11

Most of the CTC detection methods currently available come with some drawbacks, such as low overall sensitivity, the ability to capture CTCs that are positive for EpCAM or CK but not those that are negative, and the inability to perform phenotypic and genotypic characterization of CTCs following their enumeration.3, 12, 13 Thus, microfluidic devices for CTC detection have been developed that not only enhance the sensitivity of detection but also allow characterization of intact CTCs. These devices can capture viable CTCs directly from whole blood in a single step without the predilution, prelabeling, or filtration used in other techniques, all of which can result in substantial cell loss prior to detection.

The CTC chip, developed by Nagrath et al, is a microfluidic device consisting of an array of 78,000 microposts coated with anti-EpCAM antibodies on a silica chip, which has the ability to capture viable CTCs at a purity level of ∼50%, with up to 106 times enrichment.14 Another microfluidic device is the OncoCEE (Cell Enrichment and Extraction; Biocept, San Diego, Calif) platform; this device has about 9000 variably sized and randomly positioned microposts within microchannels for efficient and sensitive capturing of CTCs. The microposts are coated with streptavidin and can capture EpCAM-positive as well as some EpCAM-negative cells. This microfluidic device can also perform phenotypic and genotypic characterization of both CK-positive and CK-negative CTCs.15

A few studies have compared the sensitivities of the different detection methods. Andreopoulou et al compared a fluorescence-based microscopic assay (CellSearch) with a multiplex RT-PCR assay (AdnaTest) in 55 patients who had metastatic breast cancer and found that latter test results were positive in 53% of patients but the former test results were positive in 47% or 36% of patients, using >2 cells or >5 cells in 7.5 mL of blood, respectively, as the cutoff to designate a positive result.8 It is notable that although the RT-PCR assay was positive in 9 cases, the fluorescence-based microscopic assay was unable to detect CTCs in these cases, and the latter test was positive in 6 cases in which the former was negative. The multiplex RT-PCR assay (AdnaTest) was therefore found to be more sensitive than the fluorescence-based microscopic assay (CellSearch), but the 2 tests could be used together to improve the rate of CTC detection.

Van der Auwera et al compared CellSearch with AdnaTest and CK19/mammaglobin RT-PCR in 76 patients with metastatic breast cancer and found that RT-PCR was the most sensitive technique.10 Punnoose et al compared CellSearch with 2 commercially available microfluidic devices, OncoCEE (Biocept, San Diego, Calif) and On-Q-ity (On-Q-ity, Boston, Mass), using specimens of blood obtained directly from patients and blood samples with cell lines spiked into them.13 These authors found that CellSearch and the microfluidic devices were comparable for capturing EpCAM-expressing cells, but that efficiency fell below 50% for recovery of EpCAM-negative CTCs with both CellSearch and the microfluidic devices.

The potential clinical applications of CTC detection are of increasing interest. The presence of CTCs might be an early indicator of clinically occult metastatic disease, residual tumor burden that persists after systemic therapy, or progressive disease. The presence or absence of these cells might aid in risk stratification for patients with early or advanced breast cancer. Serial determination of CTC levels might also be a way to monitor the efficacy of different therapeutic regimens. With the aid of a noninvasive blood test, CTCs could be used as biomarkers to help select patients for targeted therapy, and repeated testing throughout the course of treatment could aid in altering therapy as needed in real time, based on changing gene expression profiles of CTCs.

The majority of the reported clinical trials related to CTCs used (CellSearch) for enumeration of CTCs in patients with metastatic breast cancer. Cristofanilli et al reported the results of the first prospective clinical trial of CTCs in patients with metastatic breast cancer.7 The fluorescence-based microscopic assay was used to monitor CTC levels in 177 patients with measurable disease before the patients were to start a new line of treatment and at the first follow-up visit. A cutoff of 5 CTCs in 7.5 mL of peripheral blood stratified the patients into 2 distinct groups; the progression-free and overall survival rates of patients who had <5 CTCs were significantly higher than those of patients who had ≥5 CTCs as detected by the assay.

Subsequent to this pivotal study, Hayes et al showed that detection of elevated CTCs at any time during the clinical course in patients with metastatic breast cancer was a harbinger of impending progression and that CTC levels may be an objective and accurate determinant of disease progression and death.16 The prognostic value of detecting CTCs in the blood of patients with metastatic breast cancer was further validated by subsequent prospective studies and has been shown to be independent of tumor burden, time to metastasis, site of metastasis, type and line of therapy, and phenotype of the primary tumor.17, 18 Studies comparing levels of CTCs with imaging findings as prognostic tools concluded that the number of CTCs in the blood was a better and earlier indicator of disease progression than were findings from traditional imaging modalities such as computed tomography and magnetic resonance imaging.19

Most clinical studies related to CTCs include patients with metastatic breast cancer; few studies have explored the status of CTCs in early-stage breast cancer, and those that did used fluorescence-based microscopic assay (CellSearch) and RT-PCR, which detected CTCs in 10% to 44% of this patient population.20, 22 However, the prognostic relevance and clinical implications of detecting CTCs in the blood of patients with early-stage breast cancer needs further investigation in prospective clinical trials.

Although most studies have focused on enumeration of CTCs as a potential prognostic tool in breast cancer, there are emerging reports in recent years showing that further characterization of the CTCs could be performed to obtain the status of protein overexpression or gene amplification of HER2. Evaluation of HER2 status using CTCs in patients with breast cancer could potentially play a role in monitoring trastuzumab therapy or in the selection of patients for trastuzumab therapy if the results of tests evaluating HER2 status differ between the primary tumor and blood CTCs. The reason for the disparity in HER2 status is not entirely clear but is thought to involve either clonal selection or genomic instability of the CTCs, resulting in acquisition of HER2 gene amplification.

Various techniques have been used to determine HER2 status in CTCs, including immunofluorescent staining for evaluating HER2 protein overexpression and FISH and PCR for evaluating HER2 gene amplification.23-28 Analysis of CTCs for HER2 expression in 2 large, multi-institutional studies by Fehm et al and Riethdorf et al was performed by immunofluorescent staining for HER2 protein overexpression, using the CellSearch assay with fluorescein isothiocyanate–labeled anti-HER2 antibodies (CellSearch tumor phenotyping reagent HER2; Veridex Inc.).24, 25 Patients were characterized as HER2-positive if at least 5 CTCs were detected and at least 1 showed strong immunofluorescence with the HER2-specific antibody. Fehm et al also used AdnaTest to determine HER2 status in CTCs, which were considered HER2-positive if a PCR fragment of the HER2 transcript showed a peak concentration of >15 ng/μL. Determination of HER2 gene amplification in CTCs by FISH was performed following enrichment of peripheral blood for EpCAM-expressing tumor cells using immunomagnetic separation. Alternatively, HER2 status can be determined in intact CTCs within the microchannels by FISH using the OncoCEE microfluidic device.

Studies testing HER2 status of CTCs using immunofluorescent staining, RT-PCR, or FISH in patients with locally advanced and metastatic breast cancer have found HER2-positive CTCs in 35% to 47% of this patient population.15, 23-28 In these studies, fewer HER2-positive CTCs were detected by FISH than by CellSearch or PCR, and AdnaTest detected the highest number of patients with HER2 mRNA overexpression in CTCs. On the basis of these and other reports, it is clear that although HER2-positive primary tumors may be associated with CTCs, the CTCs may fail to show any evidence of HER2 expression, and HER2-positive CTCs may be encountered in patients with HER2-negative primary tumors.

The prognostic information to be gained from CTCs and its potential use in the management of patients with metastatic breast cancer is evident based on prospective clinical studies, although the clinical value of detecting CTCs in early-stage breast cancer still needs validation in prospective clinical trials. However, despite the evident usefulness of CTC analysis in breast cancer and the availability of technology to perform the analysis, CTC analysis is not routinely used in a clinical setting, either as a prognostic indicator or for determining personalized therapy in patients with metastatic breast cancer. There are several issues that preclude the incorporation of CTC analysis into routine clinical practice. The most important and problematic issue is that although several CTC detection and analysis products are available, many of them are not standardized, have not been compared with each other, and have not been tested in large, prospective clinical trials. The number of detected cells in each patient and the proportion of patients in whom these cells are detected can vary depending on which platform is used for the testing. The EpCAM- and CK-based systems for enrichment and detection may not adequately detect EpCAM- and CK-negative CTCs, including those that have undergone epithelial mesenchymal transformation, and they may not always detect CTCs in patients with certain subtypes of breast cancer.12

The American Society of Clinical Oncology does not include the evaluation of CTCs in its recommendations on the use of tumor markers in patients with early and advanced-stage breast cancer, citing the need for further validation of the clinical use of the tests.29 There are a few ongoing clinical trials to evaluate the role of CTCs in the management of patients with breast cancer (listed on the National Cancer Institute Web site under Clinical Trials). The primary objectives of these trials include the following: use of CTCs to guide chemotherapy for metastatic breast cancer, correlation of levels of CTCs with clinical response to ixabepilone in patients with significant residual breast cancer, characterization of CTCs to direct preoperative and systemic therapy in patients with locally advanced or metastatic stage IV breast cancer, clinical validation of a microfluidic device for isolation and molecular characterization of CTCs, and phase 3 randomized study of treatment decision-making based on levels of CTCs in women with metastatic breast cancer undergoing chemotherapy (SWOG-S0500).30 Another ongoing prospective trial seeks to ascertain the percentage of patients with HER2-positive and HER2-negative metastatic breast cancer who have HER2-positive CTCs in their peripheral blood. The results of these trials will be important in helping us to understand the value of CTCs in clinical practice.

In addition, the value of newer and more sensitive technology (eg, microfluidic devices) for enumeration and phenotypic and genotypic characterization of EpCAM- and CK-positive and EpCAM- and CK-negative CTCs needs to be tested in large, prospective clinical trials. Likewise, the potential of CTC analysis as a tool to aid in the personalized care of patients with breast cancer needs to be evaluated in prospective clinical trials using sensitive, standardized, and reproducible CTC analysis technology to justify its incorporation into routine clinical practice.


No specific funding was disclosed.


Dr. Krishnamurthy does not have any conflicts of interest to disclose. She has not received financial compensation from the makers of CellSearch and has not a consultant or speaker for any for the companies referred to in this manuscript. Her participation in these multidisciplinary studies is purely as a scientific investigator.