Journal of Cancer Research and Therapeutics

: 2013  |  Volume : 9  |  Issue : 1  |  Page : 29--37

The prognostic value of circulating tumor cells lacking cytokeratins in metastatic breast cancer patients

Lina Zhao, Pengfei Li, Fude Li, Yongkun Yang, Na Liu, Li Cai 
 Department of Medicine, the Third Affiliated Hospital of Harbin Medical University, Harbin, Heilongjiang Province, China

Correspondence Address:
Li Cai
Department of Medicine, the Third Affiliated Hospital of Harbin Medical University, Haping Road 150 of Nangang District, Harbin, Heilongjiang Province


Aim of Study: In this study, we detected epithelial circulating tumor cells (CTCs) that expressed cytokeratins and potential circulating tumor cells (pCTCs) that had lost expression of cytokeratins in metastatic breast cancer (MBC) patients. The aim of the study was to evaluate the prognostic significance of these two kinds of CTCs in MBC patients. Materials and Methods: We detected CTCs and pCTCs from 66 MBC patients using MACS immunomagnetic enrichment technology combined with immunocytochemistry (ICC). A cutoff score of 5 CTCs (or pCTCs) was set as a benchmark for prognosis in patients. Progression free survival (PFS) was calculated and analyzed during the following 24 months. Results: We evaluated the sensitivity of this method and recovery rates of the CTCs by spiking experiments. The loss number of tumor cells by our method was 0-15. The population with fewer than 5 CTCs showed significantly higher PFS than the group with 5 or more CTCs. The difference in PFS between the patients with 5 or more pCTCs and those with fewer than 5 pCTCs was statistically significant. The presence of these pCTCs more accurately predicted poor prognosis than the CTCs that express cytokeratins. Conclusions: There is a subset of CTCs that lose epithelial markers such as pCTCs. Due to the heterogeneity of the expression of epithelial antigens in CTCs, different subtypes of CTCs exist. Independently of CTCs, the groups of patients with pCTCs had poorer prognoses.

How to cite this article:
Zhao L, Li P, Li F, Yang Y, Liu N, Cai L. The prognostic value of circulating tumor cells lacking cytokeratins in metastatic breast cancer patients.J Can Res Ther 2013;9:29-37

How to cite this URL:
Zhao L, Li P, Li F, Yang Y, Liu N, Cai L. The prognostic value of circulating tumor cells lacking cytokeratins in metastatic breast cancer patients. J Can Res Ther [serial online] 2013 [cited 2020 Feb 29 ];9:29-37
Available from:

Full Text


Distant metastasis usually portends progression of the disease and a poor prognosis for cancer patients. The emergence of circulating tumor cells (CTCs) is a pivotal step in cancer metastasis via the bloodstream. [1] In the past decade, a number of CTCs have been established to be clinically relevant for progression-free survival (PFS) and overall survival (OS) of breast cancer patients. [2],[3] The prognostic and predictive value of CTCs has been confirmed in many studies. [4],[5],[6],[7] Due to the difficulty of detecting those rare cells, studies have been enabled by the development of CTC enrichment and detection technologies.

Recently, molecular characteristics of CTCs have been shown to relate to the prognosis of the patients and to provide important prognostic and biological information about the cancer. [8],[9] Data have accumulated that show the existence of different molecular characteristics or subtypes of CTCs in breast cancer patients. [8],[10] Studies found epithelial cancer metastasis begins with the epithelial-mesenchymal transition (EMT). Subsets of CTCs are believed to have undergone EMT and lack or express low levels of epithelial markers, such as epithelial cell adhesion molecule (EpCAM) and cytokeratins (CKs). Till date however, most technologies for CTC detection, including the FDA-approved CellSearch system (USA), only detect epithelial CTCs and may miss clinically relevant subpopulations of CTCs. The EpCAM+ CK+ CD45-DAPI+ cells were defined as CTCs by the CellSearch system and CTCs with these characteristics are detected in about 60% of metastatic breast cancers (MBCs). [11] It is unclear why the detection rate is not 100%. With differences in the sensitivity of various detection methods, some CTCs that do not have anyone of those characteristics might be missed. These CTC phenotypes that lack expression of epithelial markers are perhaps more important to cancer metastasis.

This was a prospective, observational study in which MBC patients were evaluated for the presence of potential CTCs (pCTCs) that do not express cytokeratin in peripheral blood of patients with breast cancer. We performed the trial to analyze the molecular and morphological characteristics of epithelial and potential CTCs in breast cancer patients. We correlated clinical outcomes with different CTC levels in advanced breast cancer patients to explore the prognostic value of CTCs and potential CTCs. We also wanted to determine if losing expression of cytokeratins in CTCs predicts a poor prognosis in MBC patients.

 Materials and Methods

We used cultured MCF-7 cells that served as a positive control and to evaluate the sensitivity of the method and recovery rates of the CTCs by spiking experiments. MCF-7, a breast cancer cell line, was obtained from the American Type Culture Collection (ATCC). Cells were cultured in RPMI 1640 (Hyclone, UT, USA) supplemented with 10% heat-inactivated FCS (Hyclone, UT, USA), 100 units/ml penicillin, and 100μg/ml streptomycin (Invitrogen). Cells were grown at 37°C in 5% CO2. To accurately estimate tumor cell recovery, the cells were counted by Neubauer Chamber. Then we spiked a definite number of cells (10, 20, 50 and 100 MCF-7 cells) in 10 mL of peripheral blood of healthy donors. Every number of spiking experiment was repeated ten times to test the reproducibility of the cell recovery rate. We used 10 mL of peripheral blood from 10 healthy volunteers as the negative controls. The healthy female volunteers were medical workers of our hospital and everyonedonated successively total 50 mL of peripheral blood.

We selected 66 MBC patients in our hospital as subjects for this study. The recruitment period was between January 2008 and September 2009, and the follow-up period was 24 months after the blood sample extraction. The inclusion criteria were as follows: 1. All patients provided signed informed consent to donate blood samples for the study; 2. The patients underwent histopathologic confirmation of the diagnosis, had been newly diagnosed with overt metastatic disease, and had not yet started a new treatment; Patients were required to have sufficient clinical and radiologic evidence of metastatic breast cancer with either measurable or evaluable disease; 3. All participants had adequate communication and comprehension ability. The study was approved by the ethics review boards of our institution.

We drew 10 mL of blood by the cubital vein puncture from 66 MBC patients and 10 healthy female volunteers for CTC detection and morphological analysis [Figure 1]. To avoid contamination with skin cells or vascular endothelial cells, 5mL of blood was discarded or used for other routine laboratory examinations before the study samples were taken. Blood samples were maintained in EDTA (ethylenediaminetetraacetic acid) tubes at 4°C and processed within 4 hours after the blood was drawn.{Figure 1}

Volumes of 10 mL of the blood samples and 90 mL of 1× Erythrocyte Lysing Solution (Cellchip, Beijing) were mixed and kept on ice for 10-15minutes until pellucid. The mixture was then centrifuged at 400 × g for 5 minutes. The supernatant was gently decanted. The cells were rinsed with PBS (phosphate buffered saline) twice. The cells were suspended in 500μL buffer (PBS supplemented with 0.5% BSA (bull serum albumin) and 2mM EDTA)and prepared for CTC isolation.

The CD326 (EpCAM) Tumor Cell Enrichment and Detection kit including EpCAM-MicroBeads, FcR Blocking Reagent, the anti-CK mAb cocktail (CK3-3E4 and CK3-11D5) [Figure 2], MACS MS column and MACS Separator was purchased from Miltenyi Biotec (Germany). We added 100μL of FcR Blocking Reagent and 100μL of Anti-HEA (EpCAM) MicroBeads (MicroBeads conjugated to monoclonal anti-human epithelial antigen antibody HEA-125) per 5×10 7 total cells in the cell suspension. The cell suspension was mixed well and incubated for 30minutes at 4-8°C. The cell suspension was applied to the MACS MS column in the magnetic field of a MACS Separator [Figure 3]. The magnetically labeled EpCAM positive cells were retained on the column. After removal of the column from the magnetic field, the magnetically retained EpCAM positive cells were eluted and suspended in 500μL buffer for ICC [Figure 4].{Figure 2}{Figure 3}{Figure 4}

We added 500μL of Inside Fix (contains 3.7% formaldehyde) to the eluted EpCAM-positive cells and incubated for 20minutes at room temperature. We diluted the anti-CK mAb cocktail (1:50) that recognizes CK7, 8 and 18 in Inside Perm. We added 100μL of diluted antibody to the cell suspension and incubated the suspension for 10minutes at room temperature. After magnetic enrichment and intracellular staining, the cells were spun onto a glass slide using a cytocentrifuge. The slides were air-dried for at least 2 hours at room temperature and initially dyed with Fast Red TR/Naphtol AS-MX substrate solution [Figure 5]. Then, the cells were counterstainedfor 1minute in filtered Meyer's hemalum solution in a staining trough.{Figure 5}

CTCs were evaluated by two independent cell morphology observers (LF, YY). Both observers were blinded to laboratory and clinical data. Morphological analysis was performed on all specimens using a light microscope. Images were taken at 100 × magnification using an Olympus BX51 microscope (Olympus, Japan) linked with image analysis software (Maisiqi Beijing, China). Inconclusive results were discussed, and a final consensus was reached.Cells were classified as CTCs when positively stained for cytokeratins, and pCTCs were distinguished from white blood cells by virtue of their larger size and distinct nuclear morphology. The results were expressed as number of CTCs per 10mL of whole blood. CTCs and pCTCs were enumerated separately in every sample.

All patients received therapy according to the institutional and international guidelines starting after blood sample extraction and CTC detection. Patientfollow-up was carried out according to the standard hospital protocol established by the breast unit of our hospitaland the disease progression was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST). Subsequently, clinical follow-up visits were scheduled every 2 months during the following 24 months. These visits included a physical examination in which particular attention was placed on the evaluation of lesions. Before starting a new treatment, patients underwent an evaluation of the metastatic sites by ultrasound, X-ray or computer tomography. Re-evaluations of the disease status were done by the same techniques in cases of suspected relapse or progress. Curative effects were graded according to the RECIST criteria: An absence of clinical evidence of tumor was classified as a complete response (CR); at least 30% reduction in the product of the two maximum perpendicular diameters of the tumor was classified as a partial response (PR); at least 20% increasing in size or the appearance of one or more new lesions was classified as clinically progressive disease (PD); and neither sufficient shrinkage to qualify for PR nor sufficient increase to qualify for PD was classified as stable disease (SD).

All of the data was assessed by χ2 or Fisher's exact test as appropriate.A cutoff score of 5 CTCs was set as a benchmark for prognosis in patients. Progression free survival (PFS) was calculated from the date of CTC measurement to the date of clinical disease progression. The last follow-up was estimated using the Kaplan-Meier product limit method and different prognostic groups were compared using the Log rank test. All statistical tests were two-sided, and Pvalues <0.05 were considered statistically significant. All statistical calculations were performed using the SPSS 16.0 statistical software (Chicago, Illinois, USA).


We used cultured MCF-7 cells that served as a positive control. More than 90 percent of the MCF-7 cells were dyed red in ICC by Fast Red TR/Naphtol AS-MX substrate solution [Figure 6]. This demonstrated that most MCF-7 cells expressing cytokeratins, which are markers of epithelial tumor cells. Besides this staining, MCF-7 cells showed a standard morphology typical of tumor cells.{Figure 6}

In addition, in spiking experiment, analysis of serial quantity of MCF-7 cells in 10 mL normal women blood demonstrated the reliability and sensitivity of the method. The number and median recovery rates of spiked tumor cells are described in [Table 1]. We found the detection rate of spiking tumor cells elevate along with the amount increasing of tumor cells. When spiked tumor cells were 10, the rate of recovery was 62%. When the number of tumor cells reached 20, the rate of recovery was 73%. And when the number of tumor cells reached 50 and 100, the rate of recovery was 86.8% and 89.9%, respectively. These results declare that, no matter the quantity of spiked cells (≤100), the loss quantity of tumor cells by this method is no more than 15.{Table 1}

Specificity of the experiment method was tested by analyzing 10 mL of peripheral blood from 10 healthy female volunteers. We did neither detect CTCs nor pCTCs from them with this method.

A total of 66 patients (age 42-74, median age 56) were studied. The clinicopathological information of the patients with metastatic breast cancer is listed in [Table 2]. Hormone receptor positivity was detected in 35 patients (53%), and HER2 positivity was detected in 29 patients (43.9%). A total of 62 patients received chemotherapy (93.9%) and 4 patients received hormonal therapy (6.1%); within this total, 12 patients (18.2%) received local radiotherapy and 14 patients (21.2%) received Trastuzumab injections aspart of a sequential combination of both treatments.{Table 2}

CTCs captured in our assay were assessed by two appropriately trained morphologists using a standard protocol. CTCs were identified using immunophenotype and typical cytomorphology including positive cytokeratin staining, large or irregular nuclei, and a high nuclear to cytoplasmic ratio [Figure 7]. The size of CTCs was about 12-20μm, and the cytokeratins were red, granular and distributed in the cytoplasm. The nuclei were partially covered by cytokeratin staining. In addition, we observed a group of cells in some samples with a profile similar to CTCs, but there were no red, granular CKs in their cytoplasm [Figure 8] a and b. The cells in these groups were larger than white blood cells and had distinct cytomorphology.In order to observe the morphology of leukocytes and better differentiate the potential tumor cells and leukocytes, we used the same method to stain the peripheral blood smear. The representative white blood cells were present in [Figure 9]a, b, c. We hypothesized that these cells were potential CTCs (pCTCs) and that they were not stained because they had lost expression of cytokeratins. Furthermore, cytokeratins are heterogeneously expressed in the CTCs of MBC patients. Strong expression is seen in some CTCs, weak expression in others, and no expression in certain CTCs (pCTCs).{Figure 7}{Figure 8}{Figure 9}

Among the 66 blood samples of MBC patients, 42/66 (63.6%) contained CTCs (range from 1 to 128) and 34/66 (51.5%) contained pCTCs (range from 1 to 132). The distribution of the numbers of CTCs in the 66 patients with MBC is listed in [Table 3]. Overall, 30/66 (45.5%) samples contained 5 or more CTCs, and 23/66 (34.8%) samples contained 5 or morepCTCs. CTCs and pCTCs were detected in 25/66(37.9%) samples, and no CTCs were found in 12/66(18.2%) samples.{Table 3}

We used the RECIST criteria to assess progression of disease and the PFS as outcome measure. Progression free survival (PFS) was calculated from the date of CTC measurement to the date of clinical disease progression. At least 20% increase in size of old lesions or the appearance of one or more new lesions was classified as clinically progressive disease. The population with fewer than 5 CTCs showed significantly higher PFS than the group with 5 or more CTCs. The median PFS of the group containing 5 or more CTCs was 15 months vs. 21 months for the group with fewer than 5 CTCs [Figure 10]. This difference was found statistically significant (P=0.003). Independently of definite CTCs, the median PFS of patients with 5 or more pCTCs was 13 months vs. 21 months for those with fewer than 5 pCTCs [Figure 11]. This difference was statistically significant (P < 0.001).{Figure 10}{Figure 11}

According to the different types and different numbers of CTCs, we had 4 subgroups of patients: 1. 20/66 (30.3%) had 5 or more CTCs and fewer than 5 pCTCs; 2. 9/66 (13.6%) had fewer than 5 CTCs and 5 or more pCTCs; 3. 25/66 (37.9%) had 5 or more CTCs and 5 or more pCTCs; and 4. 12/66 (18.2%) had fewer than 5 CTCs and fewer than 5 pCTCs. The median PFS of the subgroups were as follows: 16, 13, 13 and 23 months, respectively [Figure 12]. The difference in PFS between the patients with 5 or more pCTCs and those with fewer than 5 pCTCs was statistically significant (P=0.002).{Figure 12}

At the time of the last follow-up, 4 (6.1%) patients had died.


CTCs are tumor cells circulating in the peripheral blood of patients that are shed either from the primary tumor or its metastases. They are relatively scarce but imply the spread of a tumor. CTCs have been shown to be a powerful prognostic biomarker in many types of cancer, including breast cancer. [12] In our investigations of CTCs, we faced two challenges: Identifying these rare cells and confirming their metastatic potential. Large differences in the detection rates of CTCs (30-100%) have been reported in breast cancer. [13],[14],[15] Besides the use of different technologies and patient populations, these discordant results could be due to different definition of CTCs. Even using the same technology, the application of different criteria for the definition of CTCs may result in different detection rates.Simplex kind of markers-based methods may miss different subpopulation of CTCs. This might be one reason wherefore the current enrichment and detection technologies detect CTCs in only up to 60%. [11] Our spiking results of the cell lines suggest that recovery rate of our method with different cell numbers was up to 62%-89.9%. In other words, in the mean 0-15 spiked tumor cells were lost during the enrichment and detection procedure. Specificity of the experiment method was tested at the same time. We did neither detect CTCs nor pCTCs with this method in peripheral blood of healthy female volunteers.

In our assay, we used the standard Magnetic Activated Cell Sorting system (MACS, Miltenyi Biotec GmbH, Germany) to enrich CTCs. MACS is a dedicated instrument that captures cells by immunomagnetic labeling with microbeads. Magnetic beads linked to an anti- EpCAM antibody are available for the selection of EpCAM+ tumor cells. EpCAM is a cell surface molecule and is highly expressed in most epithelial carcinomas. However, there is evidence that epithelial markers expression of CTCs is down regulated, [16] suggesting that EpCAM negative or/and CKs negative tumor cells circulate in peripheral blood. The process of enrich CTCs would inevitably lose some EpCAM negative CTCs. But, searching for CTCs in millions of blood cells without enrichment is highly difficult and ineffective. With the use of immunomagnetic enrichment the detection of CTCs is substantially easier and more efficient, [17] even at the cost of cell loss.

After the magnetic enrichment step, immunostaining was performed using the anti-CK mAb cocktail (CK3-3E4 and CK3-11D5) to detect cytokeratin-positive epithelial tumor cells. The CK3-3E4 and CK3-11D5 antibody cocktail is directed against common cytokeratin epitopes including the CK heterodimers 7, 8 and 18. The criteria for detection of CTCs were based on the recommendations of the European ISHAGE Working group for standardization of tumor cell detection and the consensus statements. [18],[19] CTCs were detected by morphological criteria in addition to typical immunocytologic staining. In our study population, 63.6% of MBC patients had detectable CTCs. But when we captured epithelial CTCs with cytokeratin expression, we found another group of cells under the microscope. They were not stained, and they were similar to CTCs and morphologically different from any blood cells. The size of these cells was about 12-20 μm and were larger than white blood cells. They had large or irregular nuclei, and a high nuclear to cytoplasmic ratio, but there were not red, granular cytokeratin staining in their cytoplasm. It is remarkable that the emergence of those cells was not occasional because of their quantity. We were interested in the identity of these cells. Researchers have realized that the traditional methods might not detect all kinds of CTCs. [20] It is likely that there are some CTCs that do not fit within the detection conditions, such as EpCAM negative CTCs or CKs negative CTCs.

We supposed the cells that we have observed might be a different CTC population. We have reasons to support this hypothesis. Firstly, we only used breast cancer cell line MCF-7 as positive control of CTCs because we didn't estimate the existence of the pCTCs in advance. Moreover, in process of assay, we found that more than 90 percent of the MCF-7 cells were dyed red. This demonstrated that most but not totals MCF-7 cells expressing cytokeratins. In fact, 10 percent of the MCF-7 cells lost CKs expression and didn't express red granules in their cytoplasm. The morphology characteristics of pCTCs were consistent to this 10 percent of the MCF-7 cells. Secondly, tumors have inherent genetic and phenotypic heterogeneity. [21] We cannot classify CTCs using a unified standard. Primary breast cancer tissue has a heterogeneous CKs expression. Thirdly, tumor cells will change themselves to accommodate a new environment. [22],[23] For example, epithelial tumor cells can change to mesenchymal tumor cells after activation of an epithelial-to-mesenchymal transition (EMT) program. [24] In this process, epithelial cells that have undergone EMT have enhanced migratory and invasive potential and changed to Mesenchymal CTCs. [25] Mesenchymal CTCs have a greater ability to resist attacks from the immune system and chemotherapy than epithelial CTCs. [26] They can arrive in distant organs, continue to grow and form metastases. EMT causes cells to lose epithelial markers and gain mesenchymal markers.EMT in CTCs not only promotes cell invasion and CTC generation but is also involved in therapy resistance and survival of CTCs in the blood. [27],[28] This theory can answer questions that have been raised recently about how cytokeratin-negative CTCs could predict a poor prognosis. The existence of CTCs that have lost expression of cytokeratins is reasonable. In our study, the cells in which the cytoplasm was dyed red were defined as CTCs, and the cells that were not dyed red but were consistent with tumor cells in the morphological analysis were identified as potential CTCs (pCTCs). The pCTCs that do not express CKs might be mesenchymal CTCs or other subtypes of CTCs. Thus it can be believed that CTC thresholds may be underestimated because of false negative results. The cytokeratins used for visualization of CTCs might be insufficient or some CTCs underwent phenotypic changes, such as epithelial-mesenchymal transition.

To further investigate the identity of pCTCs, we decided to explore the prognostic value of CTCs and pCTCs because true CTCs are a prognostic indicator. [29] We used the numbers of CTCs or pCTCs as a variable and analyzed the PFS of all patients in our assay. The cutoff score between positive and negative patients was set at 5 CTCs. Using this principle, we demonstrated that 5 or more CTCs were detectable in 30/66 (45.5%) patients and 5 or more pCTCs were detectable in 23/66 (34.8%) patients. We evaluated the PFS of MBC patients in a follow-up study lasting 24 months. The results demonstrated that pCTCs were most likely a subtype of CTCs because independently of CTCs, the groups of patients with pCTCs had poorer prognoses. In other words, the groups of patients with poorer prognoses are those with CTCs that do not express CKs.Thus, we may speculate that the expression of CKs in CTCs alone may not adequately predict prognosis in MBC patients.

The precision of our manual method is not as accurate as semi-automatic CellSearch system, but our method seems to be advantageous for capturing and recognizing CKs positive and CKs negative tumor cells. CTCs lacking cytokeratins would not have been counted using the CellSearch analysis because the classical definition of CTCs was not met. In frankly, neither our method nor CellSearch system did recover EpCAM negative tumor cells during the enrichment procedure. This is the strategic point that we need to improve in future experiment. We would like to mention that our study is hypothesis generating. However, this hypothesis is based on objective images and data. Next, we plan to design a prospective trial to detect mixture of epithelial and mesenchymal CTCs with a larger MBC patient cohort. Epithelial CTCs seems not to be adequately representative for the prognosis of MBC patient. The accuracy for prognostication of MBC patients depends on the accurate detection of their CTC subpopulations.

The presence of mesenchymal markers on CTCs more accurately predicted poor prognosis than the expression of cytokeratins alone. That may be the same reason that pCTCs would correlate with poor prognosis in our assay. The pCTCs were probably mesenchymal CTCs that had undergone EMT. Those mesenchymal CTCs have been associated with resistance to apoptotic signals, a property of survival in the bloodstream, resistance to chemotherapy and possibly the ability to escape immune surveillance. [30] Those disseminated CTCs have a greater potential to form secondary tumors in other organs and correlate with poor prognosis.

Our results suggest that, due to the heterogeneity of the expression of epithelial antigens in CTCs, different subtypes of CTCs exist. Using traditional assays, we may have paid more attention to epithelial CTCs and missed the most invasive group of pCTCs. Besides mesenchymal CTCs, which have undergone EMT, it is possible that other subtypes of CTCs exist. Theodoropoulos [10] showed that 35.2% of CTCs had the tumor stem cell phenotype in 30 patients with MBC. The concordance rates between the ER, PR and HER2 statuses of CTCs and the primary tumor were 29%, 25% and 53%, respectively. [31] In conclusion, the findings of our study support that there are different subtypes of CTCs in peripheral blood of MBC patients. Thus, we need to improve the detection methods and fully study biomarkers in CTCs to identify, count and characterize all CTCs. This novel information about cancer biology collected from CTCs could potentially would be used to select an appropriate targeted therapy. This could lead to the development of more effective and personalized therapies.

Our assay is the first step, and this stimulates us to research the characteristics of CTCs, as these cells are viewed as early and accessible cellular determinants of subsequent overt metastasis. Our study provides a new viewing angle for observing unrecognized populations of CTCs. In addition, such investigation would provide an opportunity to curatively approach this stage of metastasis through early intervention. Capture and study different subtypes of CTCs in patients with breast cancer would facilitate a better understanding and a more accurate description of the metastatic process.


1Kasimir-Bauer S. Circulating tumor cells as markers for cancer risk assessment and treatment monitoring. Mol Diagn Ther 2009;13:209-15.
2Cristofanilli M, Budd GT, Ellis MJ, Stopeck A, Matera J, Miller MC, et al. Circulating tumor cells, disease progression, and survival in metastatic breast cancer. N Engl J Med 2004;351:781-91.
3Hayes DF, Cristofanilli M, Budd GT, Ellis MJ, Stopeck A, Miller MC, et al. Circulating tumor cells at each follow-up time point during therapy of metastatic breast cancer patients predict progression-free and overall survival. Clin Cancer Res 2006;12:4218-24.
4Cristofanilli M, Braun S. Circulating tumor cells revisited. JAMA 2010;303:1092-3.
5Cristofanilli M, Hayes DF, Budd GT, Ellis MJ, Stopeck A, Reuben JM, et al. Circulating tumor cells: A novel prognostic factor for newly diagnosed metastatic breast cancer. J Clin Oncol 2005;23:1420-30.
6Krebs MG, Hou JM, Ward TH, Blackhall FH, Dive C. Circulating tumor cells: Their utility in cancer management and predicting outcomes. Ther Adv Med Oncol 2010;2:351-65.
7Serrano MJ, Sanchez-Rovira P, Delgado-Rodriguez M, Gaforio JJ. Detection of circulating tumor cells in the context of treatment: Prognostic value in breast cancer. Cancer Biol Ther 2009;8:671-5.
8Ignatiadis M, Kallergi G, Ntoulia M, Perraki M, Apostolaki S, Kafousi M, et al. Prognostic value of the molecular detection of circulating tumor cells using a multimarker reverse transcription-PCR assay for cytokeratin 19, mammaglobin A, and HER2 in early breast cancer. Clin Cancer Res 2008;14:2593-600.
9Lianidou ES, Mavroudis D, Sotiropoulou G, Agelaki S, Pantel K. What's new on circulating tumor cells? A meeting report. Breast Cancer Res 2010;12:307.
10Theodoropoulos PA, Polioudaki H, Agelaki S, Kallergi G, Saridaki Z, Mavroudis D, et al. Circulating tumor cells with a putative stem cell phenotype in peripheral blood of patients with breast cancer. Cancer Lett 2010;288:99-106.
11Riethdorf S, Fritsche H, Muller V, Rau T, Schindlbeck C, Rack B, et al. Detection of circulating tumor cells in peripheral blood of patients with metastatic breast cancer: A validation study of the CellSearch system. Clin Cancer Res 2007;13:920-8.
12Nagaiah G, Abraham J. Circulating tumor cells in the management of breast cancer. Clin Breast Cancer 2010;10:209-16.
13Balic M, Dandachi N, Hofmann G, Samonigg H, Loibner H, Obwaller A, et al. Comparison of two methods for enumerating circulating tumor cells in carcinoma patients. Cytometry B Clin Cytom 2005;68:25-30.
14Konigsberg R, Gneist M, Jahn-Kuch D, Pfeiler G, Hager G, Hudec M, et al.Circulating tumor cells in metastatic colorectal cancer: Efficacy and feasibility of different enrichment methods. Cancer Lett 2010;293:117-23.
15Van der AuweraI, Peeters D, BenoyI H, Elst HJ, Van Laere SJ, Prove A, et al. Circulating tumor cell detection: A direct comparison between the CellSearch System, the AdnaTest and CK-19/mammaglobin RT-PCR in patients with metastatic breast cancer. Br J Cancer 2010;102:276-84.
16Rao CG, Chianese D, Doyle GV, Miller MC, Russell T, Sanders RA Jr, et al. Expression of epithelial cell adhesion molecule in carcinoma cells present in blood and primary and metastatic tumors. Int J Oncol 2005; 27:49-57.
17Aktas B, Tewes M, Fehm T, Hauch S, Kimmig R, Kasimir-Bauer S. Stem cell and epithelial-mesenchymal transition markers are frequently over expressed in circulating tumor cells of metastatic breast cancer patients. Breast Cancer Res 2009;11:R46.
18Borgen E, Naume B, Nesland JM, Kvalheim G, Beiske K, Fodstad O, et al. Standardization of the immunocytochemical detection of cancer cells in BM and blood: I. establishment of objective criteria for the evaluation of immunostained cells. Cytotherapy 1999;1:377-88.
19Fehm T, Braun S, Muller V, Janni W, Gebauer G, Marth C, Schindlbeck C, et al. A concept for the standardized detection of disseminated tumor cells in bone marrow from patients with primary breast cancer and its clinical implementation. Cancer 2006;107:885-92.
20Konigsberg R, Obermayr E, Bises G, P feiler G, Gneist M, Wrba F, et al. Detection of Ep CAM positive and negative circulating tumor cells in metastatic breast cancer patients. Acta Oncol 2011;50:700-10.
21Pantel K, Alix-Panabieres C, Riethdorf S. Cancer micrometastases. Nat Rev Clin Oncol 2009;6:339-51.
22Kaiser J. Medicine. Cancer's circulation problem. Science 2010;327:1072-4.
23Mego M, Mani SA, Cristofanilli M. Molecular mechanisms of metastasis in breast cancer--clinical applications. Nat Rev Clin Oncol 2010;7:693-701.
24Yang J, Weinberg RA. Epithelial-mesenchymal transition: At the crossroads of development and tumor metastasis. Dev Cell 2008;14:818-29.
25Tsuji T, Ibaragi S, Hu GF. Epithelial-mesenchymal transition and cell cooperativity in metastasis. Cancer Res 2009;69:7135-9.
26Yilmaz M, ChristoforiG. EMT, The cytoskeleton, and cancer cell invasion. Cancer Metastasis Rev 2009;28:15-33.
27Bonnomet A, Brysse A, Tachsidis A, Waltham M, Thompson EW, Polette M, et al. Epithelial-to-mesenchymal transitions and circulating tumor cells. J Mammary Gland Biol Neoplasia 2010;15:261-73.
28De Wever O, Pauwels P, De Craene B, Sabbah M, Emami S, Redeuilh G, et al. Molecular and pathological signatures of epithelial-mesenchymal transitions at the cancer invasion front. Histochem Cell Biol 2008;130:481-94.
29Botteri E, Sandri MT, Bagnardi V, Munzone E, Zorzino L, Rotmensz N, et al. Modeling the relationship between circulating tumor cells number and prognosis of metastatic breast cancer. Breast Cancer Res Treat 2010;122:211-7.
30Creighton CJ, LiX, Landis M, Dixon JM, Neumeister VM, Sjolund A, et al. Residual breast cancers after conventional therapy display mesenchymal as well as tumor-initiating features. Proc Natl Acad Sci U S A 2009;106:13820-5.
31Fehm T, Hoffmann O, Aktas B, Becker S, Solomayer EF, Wallwiener D, et al. Detection and characterization of circulating tumor cells in blood of primary breast cancer patients by RT-PCR and comparison to status of bone marrow disseminated cells. Breast Cancer Res 2009;11:R59.