Purpose: To investigate the associations between baseline and posttreatment circulating tumor cell (CTC) gene expression and outcome of patients enrolled in four North Central Cancer Treatment Group metastatic breast cancer (MBC) trials in which specimens were shipped (at 4°C) from community-based sites to a reference laboratory (Mayo Clinic, Rochester, MN).

Experimental Design: Blood was collected at treating sites from MBC patients before (baseline), during, and at the end of treatment with erlotinib + gemcitabine (N0234), sorafenib (N0336), irinotecan + cetuximab (N0436), or paclitaxel-poliglumex + capecitabine (N0437). CTCs from 10 mL of EDTA blood were enriched with CD45 depletion, 24 to 30 hours postblood collection. Reverse transcription/quantitative PCR was used to determine cytokeratin-19 (CK19) and mammaglobin (MGB1) mRNA levels in CTCs from up to 13 (N0234), 16 (N0336), 18 (N0436), and 39 (N0437) patients. The gene expressions were normalized to β2-microglobulin and calibrated to healthy blood using the 2–ΔΔCq algorithm; positivity was defined as 2 or more.

Results:CK19+mRNA cells were detected in 56% to 75% and MGB1+mRNA cells in 23% to 38% of 86 patients at baseline. CK19+mRNA cells were detected in 30% to 67% and MGB1+mRNA cells in 14% to 64% of 110 postbaseline serial samples. The presence of baseline CK19+mRNA cells (P = 0.01) but not MGB1+mRNA cells (P = 0.14) was significantly associated with shorter overall survival. A decrease in MGB1+mRNA levels (baseline-week 8) seemed to be associated with clinical response (P = 0.05).

Conclusions: CTC gene expression analysis conducted by a reference laboratory is feasible when blood is collected from treating sites and processed 24 to 30 hours postcollection. The presence of baseline CK19+mRNA CTCs was associated with poor prognosis; a decrease in MGB1+mRNA CTCs may help predict response to therapy of MBC patients. Clin Cancer Res; 17(22); 7183–93. ©2011 AACR.

Translational Relevance

Previous findings using circulating tumor cells (CTCs) that are isolated at the same location of collection (and typically within 4 hours of collection) have shown that the presence of cytokeratin-19 (CK19)+mRNA cells is a poor prognostic indicator for advanced breast cancer patients. However, it is difficult to translate a gene expression–based CTC test to local/community-based treating hospitals and clinics because many do not have the required instrumentation for reverse transcription/quantitative PCR (RT-qPCR) analysis. As such, the blood needs to be shipped overnight to a reference laboratory. Therefore, we examined whether gene expression in CTCs can still predict for outcome and treatment response for patients with metastatic breast cancer (MBC) when blood is collected from local/treating sites (stored at 4°C) and CTCs are isolated from EDTA blood 24 to 30 hours postcollection at a reference laboratory. We observed that the presence of baseline CK19+mRNA CTCs was associated with traditional poor prognostic indicators (≥3 metastatic sites and higher grade), a trend toward shorter progression-free survival, and significantly shorter overall survival, independent of number of metastatic sites and grade. We also observed that a decrease in mammaglobin+mRNA CTCs may help predict response to therapy of MBC patients. Our data confirm literature findings and indicate for the first time that CK19 gene expression analysis in CTCs conducted by a reference laboratory still predicts for outcome of advanced breast cancer patients when blood is collected from local sites and processed 24 to 30 hours postcollection in the context of the North Central Cancer Treatment Group treatment trials. This will allow for a wider application of CTC gene expression monitoring in clinical trials and different practice settings. Validation studies are ongoing to confirm these findings.

Breast tumors shed malignant epithelial cells into the circulation, which can cause disease relapse. As metastasis is the main cause of death in patients with solid tumors, accurately detecting and characterizing these circulating tumor cells (CTC) can positively impact the management of breast cancer patients with metastatic disease (1, 2). The clinical utility of CTCs in breast cancer has shown promise in predicting risk, tailoring treatment, response monitoring, and developing novel therapies (3–5). The CellSearch Assay (Veridex) is an immunofluorescent-based cell imaging system for enumerating CTCs (6, 7) and is Food and Drug Administration (FDA) approved for metastatic breast cancer (MBC) patient prognosis and treatment monitoring. The presence and persistence of 5 or more CTCs per 7.5 mL blood in MBC patients before and after treatment predicted for poor clinical outcome and treatment failure (6–9) has been shown to be a superior surrogate endpoint than current radiology imaging (10) and is, perhaps, superior to standard measurements of tumor burden (11).

In addition to CTC enumeration, gene expression analysis of CTCs is becoming increasingly important in determining breast cancer patient prognosis and treatment response and understanding metastatic biology. The high multiplexing capabilities associated with reverse transcription/real-time quantitative PCR (RT-qPCR) allows for CTCs to be characterized both molecularly and biologically. Specific profiles of CTCs may result in a more effective evaluation of the clinical significance of CTCs in risk prediction and treatment monitoring (12). The AdnaTest BreastCancerSelect (AdnaGen AG) is a multiplex PCR test that detects HER2, mucin 1 (MUC1), and epithelial glycoprotein (GA733-2) transcripts (13). This test has shown clinical significance of CTCs in breast cancer patients (14), but is not FDA approved and was not available in the United States at the time of our analyses. Other multiple marker assays that include cytokeratin-19 (CK19), mammaglobin (MGB1), and HER2 are being developed to effectively detect CTCs with promising clinical significance (i.e., predicts outcome and treatment response of patients) in early- and late-stage disease (15–20). Thus, characterizing CTCs in breast cancer patients may result in an important clinical tool to aid in staging, predicting prognosis, and in designing more personalized therapeutic regimens for these patients (12).

Characterization of CTCs is attractive because blood collection is a minimally invasive procedure, and blood can be shipped from the community laboratory to the research center. However, the timing between blood collection and processing may be critical for certain analytes. Special blood collection tubes with proprietary preservatives or special phlebotomy/handling requirements are necessary for the CellSearch Assay and the Adna tests. For gene expression analyses of CTCs, EDTA tubes are typically used and in most reported studies, the blood has been processed and CTCs isolated within 2 to 4 hours after collection at the same institution of blood collection. This makes it difficult to translate a gene expression–based CTC test to local/treating sites, many of which do not have access to the required instrumentation for RT-qPCR. As such, the blood needs to be shipped overnight to a reference laboratory. Therefore, we examined whether gene expression in CTCs can still predict outcome and treatment response of breast cancer patients when CTCs are isolated from EDTA blood 24 to 30 hours postcollection at a reference laboratory. We determined the mRNA levels of CK19 and MGB1 in CTCs from MBC patients enrolled in community-based, North Central Cancer Treatment Group (NCCTG) clinical trials and investigated associations between baseline and posttreatment gene expression and patient outcome.

Patients and specimens

Patients registered to 4 phase II NCCTG MBC treatment trials, erlotinib + gemcitabine (N0234; N = 59 eligible patients), sorafenib (N0336; N = 20 eligible patients), irinotecan + cetuximab (N0436; N = 19 eligible patients), and poliglumex + capecitabine (N0437; N = 48 eligible patients; http://www.ClinicalTrials.gov; Supplementary Table S1 and Supplementary Fig. S1) were eligible for this CTC study. The CTC correlative studies associated with the clinical trials were embedded in the clinical trial protocols, which were approved by participating Institutional Review Boards. The clinical trials accrued patients between 2003 and 2007, and the blood samples were collected prospectively for CTC analyses, which were conducted blinded to the clinical trials endpoints.

Usable blood samples from up to 13 (N0234), 16 (N0336), 18 (N0436), and 39 (N0437) patients were obtained at baseline (day 1 of cycle 1, before start of therapy), during (day 1 of cycle 3; 8 weeks), and at the treatment end, typically at disease progression. Although we did not prespecify the number of patients for this CTC correlative study, blood samples for CTC analysis were not collected on all patients enrolled in the clinical trials. Compliance by the local treating hospitals and clinics of blood sample collection for CTC analysis was low (22%) for the first study (N0234) in which CTC analysis was implemented in 2002. Appropriate site training improved compliance to 80% for N0336 initiated in 2003 and 85% for N0436 and N0437 initiated in 2004. A patient flow diagram describing the number of patients and blood samples (baseline and serial) for CTC analysis is provided in Supplementary Fig. S1. Comparisons of the clinicopathologic characteristics between cohort and noncohort patients with baseline and week 8 samples are provided in Supplementary Tables S2 and S3. More noncohort patients seemed to have 3 or more metastatic sites than cohort patients (P = 0.01; Supplementary Table S2).

Blood samples (∼10 mL) for CTC gene expression analysis were collected in EDTA tubes and were drawn in conjunction but after routine clinical/hematologic blood draws to minimize skin epithelial cell contamination. Blood samples were collected at the local treating hospitals and clinics, stored at 4°C, and were shipped, on ice packs (4°C), overnight to the NCCTG research base, Mayo Clinic, Rochester, MN.

Between 24 and 30 hours postblood collection, CTCs were isolated from blood samples with the RosetteSep Human CD45 Depletion Cocktail followed by density centrifugation over Ficoll-Paque according to the manufacturers' instructions (StemCell Technologies Inc.). Density centrifugation over Ficoll-Paque is necessary for the appropriate separation of rosetted red blood and unwanted cells from the mononuclear cells of interest. The enriched peripheral blood mononuclear cell (PBMNC) layer was isolated and washed thrice with PBS/2% FBS. The PBMNC pellet was resuspended with lysis binding buffer, and the cell homogenate was stored frozen at −80°C until processed for mRNA isolation.

mRNA isolation, reverse transcription, and second-strand synthesis

The procedures for mRNA isolation and cDNA synthesis have been previously described (15). In brief, mRNA was isolated from the cell lysate by the mRNA Direct Kit and Dynabeads Oligo (dT)25 according to the manufacturers' instructions (Dynal; Invitrogen). A solid phase cDNA library was generated using AMV reverse transcriptase and oligo (dT)25 priming and stored at 4°C. Gene-specific cDNA second-strand synthesis was carried out using gene-specific forward primers as previously described (15).

Quantitative real-time PCR

CK19, MGB1, and β2-microglobulin (B2M) mRNA levels were quantified by qPCR and hydrolysis probes chemistries by a BioRad iCycler IQ. The primers and probes for CK19 (NM_002276.3) were designed with Universal Probe Library from Roche Applied Science. The forward 5′-GTCATGGCCGAGCAGAAC and reverse 3′-CCGGTTCAATTCTTCAGTCC primers were obtained from Integrated DNA Technologies. The probe (GGATGCTG) was obtained from Roche Applied Science. The primers and probe for MGB1 (HSU33147) were designed with Primer Express from Applied Biosystems Inc. The forward 5′-AGAACTGCAGGGTATGGTGAGAA and reverse 3′- ACATGTATAGCAGGTTTCAACAATTGT primers and the probe- ([6-FAM]CCAACTACGGATTGCTGCAAACCACA [BHQ1a-6FAM) were obtained from Integrated DNA Technologies. The primers and probe for B2M (NM_004048) were designed with Beacon Designer from PREMIER Biosoft International. The forward 5′-CATTCCTGAAGCTGACAGCATTC and reverse 3′-CAGAAAGAGAGAGTAGCGCGAG primers and the probe ([6-FAM] TGTCTCGCT CCGTGGCCTTAGCTG [BHQ1a-6FAM]) were obtained from Integrated DNA Technologies. Specific oligonucleotides representing the respective amplicons were obtained from Integrated DNA Technologies and were used to construct standard curves for the analytes. The oligonucleotides were serially diluted in Tris-EDTA buffer and 8 to 9 working concentrations (in multiples of 10) between 10 transcript copies per 5 μL and 108 transcript copies per 5 μL for CK19 and MGB1 and between 10 transcript copies per 5 μL and 109 transcript copies per 5 μL for B2M were assayed in duplicate. The average ± SD intrarun and interrun (13 runs total for the 4 clinical trials) percent coefficient of variation (%CV) for CK19 across the oligonucleotide transcript levels were 1.58% ± 0.40% and 10.5 ± 4.64%, respectively. The average intrarun and interrun %CV for MGB1 across the oligonucleotide transcript levels were 1.28 ± 0.96 and 8.21 ± 3.59, respectively. The average (± SD) intrarun and interrun %CV for B2M across the oligonucleotide transcript levels were 0.82% ± 0.38% and 11.3% ± 6.61%, respectively. These %CVs are within the FDA guidelines that recommend %CV values of less than 15% for analytic assay precision and at least two thirds of the oligonucleotide transcript levels (standards) have %CV less than 15% (21).

Appropriate negative (e.g., water and known amount of mRNA from the MDA MB-361 cell line that contains the marker of interest reverse transcribed without AMV reverse transcriptase) and positive (e.g., specific amplicon oligonucleotides and known amount of mRNA from the MDA MB-361 cell line reverse transcribed with AMV reverse transcriptase) controls were included in each qPCR assay. In accordance to the Minimum Information for Publication of Quantitative PCR Experiments guidelines (22), the specific oligonucleotide standard curves were used to optimize amplification efficiency and to ensure that reaction efficiencies are comparable between genes of interest and the reference gene. In addition, blood samples obtained from healthy individuals were identically processed and were analyzed in each RT-qPCR run. The analytic detection limit of our model assay system is 1 in 106 cells (15), and with oligonucleotide standard curves, we can routinely detect at least 3 transcript copies per reaction. All patient samples were run in triplicate, and a sample had to have at least 2 Cq values (Cq is defined as the fractional cycle number at which the fluorescence generated by cleavage of the probe passes a fixed threshold above baseline) less than 40 to be evaluable for the respective transcript. Samples with at least 2 Cq values of 40 were considered nondetectable. The average Cq value was used as the quantitative value, and the average %CVs (± SD) of the detectable samples were 1.66 ± 0.12, 0.97 ± 0.07, and 0.48 ± 0.02 for CK19, MGB1, and B2M, respectively, across the 4 studies.

The target messages in the CTC samples were determined using a relative quantification method (22–24). The relative quantification, 2–ΔΔCq algorithm, is the fold expression relative to a calibrator (i.e., blood from healthy individuals; “normal blood”) and normalized to a reference gene (B2M). For the CK19 calibrator, of 62 blood samples tested from healthy individuals, CK19Cq less than 35 was observed in 11% of samples. For the MGB1 calibrator, of 56 samples tested from healthy individuals, MGB1Cq less than 35 was observed in less than 2% of samples. The average calibrator ΔCq value was obtained from these samples and individual ΔCqs of the calibrator were within 2 SDs of the average calibrator ΔCq value. Samples were classified as positive for a particular gene if the 2–ΔΔCq was 2.0 or more (i.e., 100% or more than what is found in healthy blood). Samples with nondetectable target messages (Cq = 40) were assigned a 2–ΔΔCq value of 0.

Statistical analysis

The primary endpoints of this analysis were to evaluate the impact of baseline and follow-up changes of CK19mRNA+ and MGB1mRNA+ CTCs on progression-free survival (PFS) and overall survival (OS) of MBC patients enrolled in 4 NCCTG phase II treatment trials. This study followed REMARK guidelines (25). Wilcoxon rank-sum tests were used to assess changes in CK19 and MGB1 gene expression values between baseline and week 8 and end of treatment between responders and nonresponders. Wilcoxon signed-rank tests were used to assess changes in CK19 and MGB1 gene expression levels between baseline and week 8 and treatment end. χ2 tests were used to determine whether CK19 and MGB1 gene expression values differed between studies. Nonstratified Cox regression models and Kaplan–Meier curves were used to determine associations between PFS and OS and CK19mRNA and MGB1mRNA positivity. Multivariate modeling methods included backward elimination, stepwise and assessing all potential subset models based on score criterion. The Kaplan–Meier method was used to estimate OS and PFS. Median follow-up was 2.3 years. For each clinical trial, clinical response rates as defined by Response Evaluation Criteria in Solid Tumors (RECIST) v1.0 criteria (26) among patients included in this study are listed in Supplementary Table S1.

Distribution of CK-19 and MGB1 relative gene expression levels

Table 1 shows that 38% and 67% of patients had CK19+mRNA and MGB1mRNA levels less than 2.0, respectively, at baseline. Of the 53 patients with CK19 2–ΔΔCq of 2 or more at baseline, the relative gene expression level ranged between 2.1 and 16,286 (mean: 569 median: 54; Supplementary Table S4). Of the 28 patients with MGB1 2–ΔΔCq of 2 or more at baseline, the relative gene expression level ranged between 2.3 and 5806 (mean: 354 median: 14; Supplementary Table S5).

Table 1.

Distribution of CK19 and MGB1 relative gene expression levels

Baseline (N = 86)Week 8 (N = 63)Treatment end (N = 47)
N (%)N (%)N (%)
CK19 (2–ΔΔCq
Negative 33 (38) 31 (49) 22 (47) 
 0 25 (29) 24 (38) 19 (40) 
 0.01–1.0 3 (3) 4 (6) 1 (2) 
 1.01–1.99 5 (6) 3 (5) 2 (4) 
Positive 53 (62) 32 (51) 25 (53) 
 2.0–9.99 15 (17) 10 (16) 9 (19) 
 >10.0 38 (44) 22 (35) 16 (34) 
MGB1 (2–ΔΔCq
Negative 58 (67) 42 (67) 32 (68) 
 0 51 (59) 36 (57) 29 (62) 
 0.01–1.0 5 (6) 5 (8) 3 (6) 
 1.01–1.99 2 (2) 1 (2) 0 (0) 
Positive 28 (33) 21 (33) 15 (32) 
 2.0–9.99 8 (9) 8 (13) 5 (11) 
 >10.0 20 (23) 13 (21) 10 (21) 
Baseline (N = 86)Week 8 (N = 63)Treatment end (N = 47)
N (%)N (%)N (%)
CK19 (2–ΔΔCq
Negative 33 (38) 31 (49) 22 (47) 
 0 25 (29) 24 (38) 19 (40) 
 0.01–1.0 3 (3) 4 (6) 1 (2) 
 1.01–1.99 5 (6) 3 (5) 2 (4) 
Positive 53 (62) 32 (51) 25 (53) 
 2.0–9.99 15 (17) 10 (16) 9 (19) 
 >10.0 38 (44) 22 (35) 16 (34) 
MGB1 (2–ΔΔCq
Negative 58 (67) 42 (67) 32 (68) 
 0 51 (59) 36 (57) 29 (62) 
 0.01–1.0 5 (6) 5 (8) 3 (6) 
 1.01–1.99 2 (2) 1 (2) 0 (0) 
Positive 28 (33) 21 (33) 15 (32) 
 2.0–9.99 8 (9) 8 (13) 5 (11) 
 >10.0 20 (23) 13 (21) 10 (21) 

Incidence of CK-19mRNA± and MGB1mRNA± CTCs

CK19mRNA+ CTCs (at 2–ΔΔCq ≥2) were detected in 55% to 75%, 44% to 60%, and 30% to 67% patients at baseline, during, and end treatment/progression, respectively (Table 1, Supplementary Fig. 2A). MGB1mRNA+ CTCs were detected in 23% to 38%, 22% to 64%, and 14% to 60% patients at baseline, during, and treatment end, respectively (Table 1, Supplementaryl Fig. 2B). No significant differences were observed between the time points across the 4 studies. At baseline, the incidence of CK19mRNA+ CTCs (62%; 95% CI: 51–72) was higher than the incidence of MGB1mRNA+ CTCs (33%; 95% CI: 23–42). Of the 28 MGB1+mRNA samples, 23 (82%) were also CK19mRNA+ (Supplementaryl Table S6), and of the 53 CK19+mRNA samples, 23 (43%) were also MGB1mRNA+. At baseline, CK19+mRNA and MGB1+mRNA incidence was significantly correlated (Supplementaryl Table S6; P = 0.007), and the relative gene expression levels between CK19 and MGB1 were weakly correlated (Spearman r2 = 0.23; P = 0.03).

Correlation between CTC positivity and clinicopathologic characteristics

The presence of CK19mRNA+ CTCs (2–ΔΔCq ≥ 2) at baseline correlated with hormone receptor positivity, more than 3 metastatic sites, and nonductal histology but was not correlated with nodal status (Table 2). The presence of MGB1mRNA+ CTCs (2–ΔΔCq ≥ 2) at baseline correlated with hormone receptor positivity (Table 2).

Table 2.

Correlation between CTC baseline positivity and clinicopathologic characteristics

CharacteristicCK19 baseline (N = 86)MGB1 baseline (N = 86)
N (%) ≥ 2PaN (%) ≥ 2P
Nodal status  0.82  0.65 
 Node negative 27/43 (63)  13/43 (30)  
 Node positive 26/43 (61)  15/43 (35)  
Histologic tumor grade (Elston/SBR) (N = 69) 0.71 (N = 69) 0.48 
 Well/intermediate 12/20 (60)  7/20 (35)  
 Poor 27/49 (55)  13/49 (27)  
Surgery (N = 70) 0.19 (N = 70) 0.12 
 Breast conserving 16/22 (73)  5/22 (23)  
 Mastectomy 27/48 (56)  20/48 (42)  
ER/PgR status  0.03  0.009 
 ER or PgR positive 32/44 (73)  20/44 (46)  
 Other 21/42 (50)  8/42 (19)  
# of metastatic sites  0.009  0.90 
 <3 25/50 (50)  16/50 (32)  
 ≥3 28/36 (78)  12/36 (33)  
Histology (N = 57) 0.004 (N = 57) 0.12 
 Ductal 23/48 (48)  14/48 (29)  
 Other 9/9 (100)  5/9 (56)  
Performance status  0.15  0.66 
 0 27/49 (55)  15/49 (31)  
 1,2 26/37 (70)  13/37 (35)  
Menopausal status  0.38  0.35 
 Pre- or <50 y 13/24 (54)  6/24 (25)  
 Post- or ≥50 y 40/62 (65)  22/62 (36)  
Age group  0.61b  0.72b 
 <40 6/10 (60)  4/10 (40)  
 40–49 10/17 (59)  5/17 (29)  
 50–59 19/31 (61)  11/31 (36)  
 60–60 14/23 (61)  6/23 (26)  
 70+ 4/5 (80)  2/5 (40)  
CharacteristicCK19 baseline (N = 86)MGB1 baseline (N = 86)
N (%) ≥ 2PaN (%) ≥ 2P
Nodal status  0.82  0.65 
 Node negative 27/43 (63)  13/43 (30)  
 Node positive 26/43 (61)  15/43 (35)  
Histologic tumor grade (Elston/SBR) (N = 69) 0.71 (N = 69) 0.48 
 Well/intermediate 12/20 (60)  7/20 (35)  
 Poor 27/49 (55)  13/49 (27)  
Surgery (N = 70) 0.19 (N = 70) 0.12 
 Breast conserving 16/22 (73)  5/22 (23)  
 Mastectomy 27/48 (56)  20/48 (42)  
ER/PgR status  0.03  0.009 
 ER or PgR positive 32/44 (73)  20/44 (46)  
 Other 21/42 (50)  8/42 (19)  
# of metastatic sites  0.009  0.90 
 <3 25/50 (50)  16/50 (32)  
 ≥3 28/36 (78)  12/36 (33)  
Histology (N = 57) 0.004 (N = 57) 0.12 
 Ductal 23/48 (48)  14/48 (29)  
 Other 9/9 (100)  5/9 (56)  
Performance status  0.15  0.66 
 0 27/49 (55)  15/49 (31)  
 1,2 26/37 (70)  13/37 (35)  
Menopausal status  0.38  0.35 
 Pre- or <50 y 13/24 (54)  6/24 (25)  
 Post- or ≥50 y 40/62 (65)  22/62 (36)  
Age group  0.61b  0.72b 
 <40 6/10 (60)  4/10 (40)  
 40–49 10/17 (59)  5/17 (29)  
 50–59 19/31 (61)  11/31 (36)  
 60–60 14/23 (61)  6/23 (26)  
 70+ 4/5 (80)  2/5 (40)  

Abbreviations: ER, estrogen receptor; PgR, progesterone receptor; SBR, Scarff-Bloom-Richardson.

aχ2 test.

bMantel–Haesnzel test.

Association between baseline CTC positivity and patient survival

Univariate analysis showed that histologic grade and performance status were independent significant predictors of PFS (Table 3). The HR for PFS was 1.37 (95% CI: 0.87–2.18; P = 0.18) for patients with CK19mRNA+ CTCs (2–ΔΔCq ≥ 2) at baseline (Table 3, Fig. 1A). PFS for patients with and without MGB1mRNA+ CTCs were similar, and there was no significant association between baseline MGB1mRNA+ positivity and PFS (Fig. 1B). A comparison across 4 subgroups based on baseline CTC gene expression positivity (e.g., CK19mRNA–/MGB1mRNA–; CK19mRNA–/MGB1mRNA+;CK19mRNA +/MGB1mRNA–;CK19mRNA+/MGB1mRNA+) showed no significant differences in PFS (Supplementaryl Table S8).

Figure 1.

Baseline CTC positivity and patient survival. A, CK19mRNA+ and PFS; B, MGB1mRNA+ and PFS; C, CK19mRNA+ and OS; D, MGB1mRNA+ and OS; E, CK19mRNA+ and MGB1mRNA+ and OS.

Figure 1.

Baseline CTC positivity and patient survival. A, CK19mRNA+ and PFS; B, MGB1mRNA+ and PFS; C, CK19mRNA+ and OS; D, MGB1mRNA+ and OS; E, CK19mRNA+ and MGB1mRNA+ and OS.

Close modal
Table 3.

Univariate analysis for PFS and OS

CovariateHR (95% CI)P
PFS 
 Age 1.00 (0.98–1.02) 1.0 
 ER (negative vs. positive) 1.48 (0.94–2.32) 0.09 
 PgR (negative vs. positive) 1.27 (0.80–2.02) 0.32 
 Histologic tumor grade Elston/SBR (poor vs. well/intermediate) 1.89 (1.05–3.40) 0.03 
 Node (positive vs. negative) 0.89 (0.57–1.38) 0.59 
 Menopausal status (<50 vs. ≥50 y) 1.12 (0.68–1.85) 0.65 
 Histology (other vs. ductal) 0.86 (0.40–1.83) 0.69 
 Performance status (1,2 vs. 0) 1.61 (1.03–2.51) 0.04 
 # Metastatic sites (≥3 vs. <3) 1.15 (0.73–1.81) 0.54 
 CK19 (≥2.0 vs. <2.0) 1.37 (0.87–2.18) 0.18 
 MGB1 (≥2.0 vs. <2.0) 1.05 (0.65–1.69) 0.84 
OS 
 Age 1.00 (0.98–1.03) 0.87 
 ER (negative vs. positive) 1.52 (0.90–2.55) 0.11 
 PgR (negative vs. positive) 1.52 (0.89–2.61) 0.12 
 Histologic tumor grade Elston/SBR (poor vs. well/intermediate) 2.28 (1.08–4.85) 0.03 
 Node (positive vs. negative) 1.12 (0.67–1.88) 0.66 
 Menopausal status (<50 vs. ≥50 y) 1.23 (0.67–2.17) 0.47 
 Histology (other vs. ductal) 1.48 (0.61–3.61) 0.38 
 Performance status (1,2 vs. 0) 1.13 90.67–1.91) 0.64 
 # Metastatic sites (≥3 vs. <3) 2.00 (1.18–3.41) 0.01 
CK19mRNA (≥2.0 vs. <2.0) 2.05 91.16–3.62) 0.01 
MGB1mRNA (≥2.0 vs. <2.0) 1.49 (0.88–2.52) 0.14 
CovariateHR (95% CI)P
PFS 
 Age 1.00 (0.98–1.02) 1.0 
 ER (negative vs. positive) 1.48 (0.94–2.32) 0.09 
 PgR (negative vs. positive) 1.27 (0.80–2.02) 0.32 
 Histologic tumor grade Elston/SBR (poor vs. well/intermediate) 1.89 (1.05–3.40) 0.03 
 Node (positive vs. negative) 0.89 (0.57–1.38) 0.59 
 Menopausal status (<50 vs. ≥50 y) 1.12 (0.68–1.85) 0.65 
 Histology (other vs. ductal) 0.86 (0.40–1.83) 0.69 
 Performance status (1,2 vs. 0) 1.61 (1.03–2.51) 0.04 
 # Metastatic sites (≥3 vs. <3) 1.15 (0.73–1.81) 0.54 
 CK19 (≥2.0 vs. <2.0) 1.37 (0.87–2.18) 0.18 
 MGB1 (≥2.0 vs. <2.0) 1.05 (0.65–1.69) 0.84 
OS 
 Age 1.00 (0.98–1.03) 0.87 
 ER (negative vs. positive) 1.52 (0.90–2.55) 0.11 
 PgR (negative vs. positive) 1.52 (0.89–2.61) 0.12 
 Histologic tumor grade Elston/SBR (poor vs. well/intermediate) 2.28 (1.08–4.85) 0.03 
 Node (positive vs. negative) 1.12 (0.67–1.88) 0.66 
 Menopausal status (<50 vs. ≥50 y) 1.23 (0.67–2.17) 0.47 
 Histology (other vs. ductal) 1.48 (0.61–3.61) 0.38 
 Performance status (1,2 vs. 0) 1.13 90.67–1.91) 0.64 
 # Metastatic sites (≥3 vs. <3) 2.00 (1.18–3.41) 0.01 
CK19mRNA (≥2.0 vs. <2.0) 2.05 91.16–3.62) 0.01 
MGB1mRNA (≥2.0 vs. <2.0) 1.49 (0.88–2.52) 0.14 

Abbreviations: SBR, Scarff-Bloom-Richardson; ER, estrogen receptor; PgR, progesterone receptor.

Univariate analyses showed that histologic grade, number of metastatic sites, and CK19mRNA+ CTCs were independent predictors for OS (Table 3). Patients with CK19mRNA+ CTCs had a significantly worse OS than patients without CK19mRNA+ CTCs at baseline (Fig. 1C). The HR for patients with CK19+mRNA CTCs was 2.05 (95% CI: 1.16–3.62; Fig. 1D). The HRs for patients with CK19+mRNA andMGB1mRNA+ CTCs and CK19+mRNA or MGB1mRNA+ CTCs were 1.85 (95% CI: 1.07–3.20; P = 0.03) and 1.91 (95% CI: 1.04–3.51; P = 0.04), respectively (Fig. 1E). In a multivariate analysis, the histologic grade and number of metastatic sites were not significant covariates in the prognostic utility of CK19mRNA+ CTCs on OS. CK19mRNA+ (P = 0.001) and estrogen receptor negativity (P = 0.007) were the only predictors of worse OS in the final multivariate model (Supplementary Table 7).

An analysis assessing the association of baseline CTC gene expression positivity with OS showed that patients with CK19mRNA+/MGB1mRNA− CTCs had a nonsignificant HR of 1.77 (P = 0.10) and patients with CK19mRNA+/MGB1mRNA+ had a significant HR of 2.44 (P = 0.01) when compared with patients with CK19mRNA−/MGB1mRNA− (Supplementaryl Table S9).

Association between baseline and week 8 CTC positivity change and patient OS

Patients who remained positive (2–ΔΔCq ≥ 2) between baseline and week 8 for CK19mRNA+ (HR: 2.67; P = 0.05) or MGB1mRNA+ (HR: 2.13; P = 0.09) CTCs seemed to have a worse OS than patients who remained negative (2–ΔΔCq < 2) for CK19mRNA or MGB1mRNA CTCs, respectively (Table 4). Similarly, the patients who were positive for CK19mRNA CTCs at baseline irrespective of status at 8 weeks had a worse OS outcome than patients who were negative at baseline irrespective of status at 8 weeks (HR = 2.7; P = 0.01). The patients who were positive for MGB1mRNA CTCs at baseline irrespective of status at 8 weeks tend to have a worse OS outcome than patients who were negative at baseline irrespective of status at 8 weeks (HR = 1.91; P = 0.07).

Table 4.

OS by change in CTC positivity between baseline and week 8

CTC changeN = 59#EventsHR (95% CI)POS
1 y2 y
CK19 
 <2.0 to <2.0 14 1.0 (—) — 85.7 61.9a 
 <2.0 to ≥2.0 10 0.63 (0.15–2.54) 0.51 87.5a 87.5a 
 ≥2.0 to <2.0 16 12 1.93 (0.71–5.22) 0.20 75.0 35.7a 
 ≥2.0 to ≥2.0 19 13 2.67 (0.99–7.15) 0.05 73.7 21.6a 
MGB1 
 <2.0 to <2.0 30 15 1.0 (—) — 80.0 53.8 
 <2.0 to ≥2.0 0.92 (0.30–2.79) 0.88 76.2a 63.5a 
 ≥2.0 to <2.0 11 1.63 (0.65–4.08) 0.29 80.8a 25.3a 
 ≥2.0 to ≥2.0 2.13 (0.90–5.06) 0.09 77.8a 25.9a 
CTC changeN = 59#EventsHR (95% CI)POS
1 y2 y
CK19 
 <2.0 to <2.0 14 1.0 (—) — 85.7 61.9a 
 <2.0 to ≥2.0 10 0.63 (0.15–2.54) 0.51 87.5a 87.5a 
 ≥2.0 to <2.0 16 12 1.93 (0.71–5.22) 0.20 75.0 35.7a 
 ≥2.0 to ≥2.0 19 13 2.67 (0.99–7.15) 0.05 73.7 21.6a 
MGB1 
 <2.0 to <2.0 30 15 1.0 (—) — 80.0 53.8 
 <2.0 to ≥2.0 0.92 (0.30–2.79) 0.88 76.2a 63.5a 
 ≥2.0 to <2.0 11 1.63 (0.65–4.08) 0.29 80.8a 25.3a 
 ≥2.0 to ≥2.0 2.13 (0.90–5.06) 0.09 77.8a 25.9a 

aLess than 10 at risk.

Association between CTC positivity and patient response

No significant correlations were observed between treatment response and CK19mRNA+ CTCs at any time point (data not shown). At week 8, 41% (17 of 42) of patients without MGB1mRNA+ CTCs had a response compared with 19% (4 of 21) of patients with MGB1mRNA+ CTCs (P = 0.09). At treatment end, 38% (12 of 32) of patients without MGB1mRNA+ CTCs had a response compared with 7% (1 of 15) of patients with MGB1mRNA+ CTCs (P = 0.03). Table 5 shows that the decrease in MGB1 mRNA levels seemed to be larger for patients who had a treatment response compared with those who did not have treatment response at week 8 (P = 0.05). However, it appears that MGB1 mRNA levels increased between baseline and treatment end in patients with a treatment response (P = 0.09). There were no significant associations between changes in CK19+mRNA levels and treatment response (data not shown).

Table 5.

Gene expression level change and treatment response

Response groupChange in CTC levels from baseline to week 8
MeanMedianRangeQuartilesP
CK19 
 Yes (N = 21) −350 −4063 to −865 25%: −1 0.25 
    75%: −204  
 No (N = 38) −75 −2 −16,272 to −11,906 25%: 2  
    75%: −13  
Mammoglobin 
 Yes (N = 21) −67 −738 to −36.0 25%: 0 0.05 
    75%: −11  
 No (N = 38) −32 −970 to −313 25%: 7  
    75%: − 0.1  
Response group Change in CTC levels from baseline to end of treatment 
 Mean Median Range Quartiles P 
CK19 
 Yes (N = 13) 2732 −3392 to −39,385 25%: 2 0.34 
    75%: −141  
 No (N = 31) −479 −15,332 to −999 25%: 10  
    75%: −8  
Mammoglobin 
 Yes (N = 13) 1176 −192 to −5,506 25%: 0 0.09 
    75%: −0.6  
 No (N = 31) −0.06 −970 to −459 25%: 9  
    75%: −0.1  
Response groupChange in CTC levels from baseline to week 8
MeanMedianRangeQuartilesP
CK19 
 Yes (N = 21) −350 −4063 to −865 25%: −1 0.25 
    75%: −204  
 No (N = 38) −75 −2 −16,272 to −11,906 25%: 2  
    75%: −13  
Mammoglobin 
 Yes (N = 21) −67 −738 to −36.0 25%: 0 0.05 
    75%: −11  
 No (N = 38) −32 −970 to −313 25%: 7  
    75%: − 0.1  
Response group Change in CTC levels from baseline to end of treatment 
 Mean Median Range Quartiles P 
CK19 
 Yes (N = 13) 2732 −3392 to −39,385 25%: 2 0.34 
    75%: −141  
 No (N = 31) −479 −15,332 to −999 25%: 10  
    75%: −8  
Mammoglobin 
 Yes (N = 13) 1176 −192 to −5,506 25%: 0 0.09 
    75%: −0.6  
 No (N = 31) −0.06 −970 to −459 25%: 9  
    75%: −0.1  

The detection of CTCs has been shown to predict for poor clinical outcome and treatment failure in MBC patients (6, 7). The biological characteristics (e.g., protein and mRNA expression) of CTCs may further improve predicting risk, monitoring response, and tailoring treatment strategies for individual patients with breast cancer (4, 5, 12). As CTC molecular characterization is becoming more relevant in personalized disease management of breast cancer patients, it would be beneficial to develop a gene expression–based test in which a commonly used blood collection tube (e.g., EDTA) can be easily used by local sites and the blood easily shipped to a reference laboratory and reliably tested for molecular analysis. This would potentially increase the accessibility of these types of tests to many patients, particularly in the United States. Therefore, we determined the feasibility of analyzing CK19 and MGB1 gene expression in CTCs when isolated from EDTA blood at a central laboratory 24 to 30 hours postcollection in context of outcome of patients enrolled in NCCTG community–based treatment trials.

We observed that CK19+mRNA was detected overall in 62% of MBC patients at baseline, a slightly higher incidence than the approximately 50% incidence of patients with 5 or more CTCs typically observed by the FDA-approved CellSearch Assay (6) and 52% incidence by the AdnaTest BreastCancer (27). In 2 individual studies (N0436 and N0437) with several patients treated in the first-line setting, we did observe similar incidences of baseline CTC positivity (55%–56%) compared with literature findings. MGB1+mRNA has been detected in 33% to 39% of patients (28, 29), which is within the 20% to 40% range of baseline incidence we observed using our immunodepletion/qRT-PCR method. A recent comparison analysis of the CellSearch Assay, Adna Test Breast Cancer Select/Detect, and CK19/MGB1 RT-qPCR assay detected CTCs in 36%, 22%, and 26% (CK19)/54% (MGB1), respectively, in MBC patients (30). Our increased sensitivity of CK-19+ CTCs compared with previous results may be due to different patient populations (first versus second/third/fourth line), cell enrichment and RNA extraction techniques, oligonucleotide primers, and PCR approaches (standard versus real time) used throughout the various studies. These technical differences for gene expression–based tests emphasize the importance of implementing standardize approaches to isolate and characterize CTCs, similar to what has been proposed to analyze disseminated tumor cells in the bone marrow (31).

The MGB1+mRNA incidence was lower than the incidence of CK19mRNA+ detection. Although the majority (87%) of the MGB1mRNA+ samples were also CK19mRNA+, 43% of CK19mRNA+ samples were MGB1mRNA+ and CK19 and MGB1 transcript levels were only weakly correlated. We previously observed a significant correlation between CK19 and MGB1 expression in patients with primary breast cancer (15). The presence of CK19mRNA+ CTCs at baseline was correlated with hormone receptor positivity, nonductal histology, more than 3 metastatic sites, but not with nodal status, and the presence of baseline MGB1mRNA+ CTCs was strongly associated with hormone receptor positivity. It is interesting that significant correlations between CTC gene expression (particularly for MGB1) and hormone receptor positivity were observed. Elevated MGB1 expression has been associated with clinical and biological features defining a less aggressive tumor phenotype in breast tumors (32) including steroid receptor–positive breast tumors (33). This may help explain why baseline MGB1 gene expression did not predict for reduced PFS or OS in our study. Our results are consistent with previous findings that showed a positive association between the presence of CTCs as detected by laser scanning cytometry and RT-qPCR and distant metastases (34–36). Although earlier results suggest that the incidence of CTCs is higher for node-positive than node-negative patients (34, 36, 37), the lack of correlation between CTCs and nodal status has been previously observed and indicates that 2 distinct routes exist for breast cancer metastases (38–40). Our incidence results and clinicopathologic correlations are consistent with previous findings indicating that our immunodepletion coupled with RT-qPCR approach produces reliable gene expression analysis of CTCs.

Importantly, we observed that the presence of CK19mRNA+ CTCs is associated with shorter OS and thus, correlates with poor prognosis. We also observed a potential trend toward a predictive significance of MGB1mRNA+ CTCs in that the decrease in MGB1+ transcript levels between baseline and 8 weeks seemed to correlate with patient response to treatment. Similarly, MGB1 transcripts have been recently shown to reflect the effect of therapy on adjuvant breast cancer patients (41). As MGB1 is more highly breast specific compared with CK19, which is used as a general epithelial marker, MGB1 may be a more appropriate measure of treatment response (e.g., tumor shrinkage) for breast cancer compared with CK19 (CK19 gene expression also could be confounded by its loss through epithelial–mesenchymal transition). However, increased response rates may not always result in improved PFS or OS (42). In our study, the baseline or the change in MGB1 gene expression did not significantly predict for patient PFS or OS. In addition, the detection rate of CK19+ or MGB1+ transcripts in blood samples was not shown to be associated with tumor progression in MBC patients (30). It is also interesting that a trend (P = 0.09) was observed between increased MGB1 mRNA levels (between baseline and treatment end) and treatment response. Our initial hypothesis was that a decrease in CTC gene expression would predict that a patient would respond to treatment. We did observe this for MGB1 at the first follow-up at 8 weeks but potentially the opposite was observed at treatment end. A possible explanation could be that initially, the treatment is rapidly clearing the tumor cells from the circulation and by treatment end the tumor is shrinking, releasing cells into the circulation at a greater rate than the cells being cleared from the circulation giving the appearance that an increase in CTC MGB1 gene expression at treatment end may be indicative of treatment response. However, it is important to note that the change in MGB1 gene expression between baseline and treatment end was nonsignificant and could be due to the limited sample size and various treatment regimens used between the different clinical trials.

Our findings, however, are consistent with previous results that showed an association between the presence of CK19+ CTCs and poor prognosis (16, 27, 43, 44) and show the validity of the immunodepletion/RT-qPCR method used in this study. In addition, it seems that patients who tested negative for CK19+CTCs at baseline, irrespective of changing to positive at follow-up time points, tend to have a better outcome than patients who tested positive for CK19+CTCs at baseline irrespective of changing to negative at follow-up time points. This is in slight contrast to the CellSearch data that showed that patients who changed from negative (<5 CTCs per 7.5mL blood) to positive (≥5 CTCs per 7.5mL blood) had worse outcome than those patients who changed from positive to negative and the outcome of those patients who turned positive was similar to those patients who remained positive for CTCs (6).

Concern does exist about the decreased gene expression levels observed in room temperature EDTA blood samples 6 hours postcollection (45). The blood samples in our studies were stored at 4°C and shipped on ice packs. This cold temperature may slow RNA degradation as our developmental work with blood from healthy individuals spiked with 1,000 cells/mL blood of MDA-MB-361 cells and stored at 4°C suggests that mRNA levels are decreased to a lesser extent when blood was maintained at 4°C (average Cq increase of 1.63 and 1.21 for CK19 and MGB1, respectively, as observed in our hands) versus a Cq increase of 2 to 4 when EDTA blood was kept at room temperature for 24 hours (45). A Cq increase of 6.29 and 5.14, for CK19 and MGB1, respectively, was observed when blood was collected in a CellSave tube maintained at room temperature (Supplementaryl Fig. 3). Although RNA degradation occurs, small transcripts are still detectable by real-time PCR, a technique that is tailored to amplify small amplicon sizes. In addition, large and small tumor cell fragments and tumor microparticles have been observed in blood from colon cancer patients using the CellSearch Assay, and these small fragments were still associated with poor clinical outcome of the patients (46). In addition, EpCam antigenicity was still present 3 days after blood collection (47). Thus, it seems that RNA may be coming from tumor cell fragments that are still clinically significant and selected cells remain viable for several days after collection. Last, our quantification method couples immunodepletion cell enrichment with RT-qPCR for which the gene expression results are calibrated to the gene expression observed in healthy blood that is identically processed (i.e., processed 24 hours postcollection) as the samples obtained from patients enrolled in NCCTG clinical trials. The negative selection has been shown to reduce background signals of HER2 and CK19 in hematopoietic cell populations to improve the specificity of RT-qPCR (48). We used normal blood as a calibrator in our quantification method to further decrease and normalize the influence of marker gene expression potentially found in hematopoietic cells as we did observe relatively elevated CK19 gene expression (e.g., Cq < 35) in 11% of normal blood samples. Molecular characterization of a single CTC by a multiplex real-time PCR also has been shown in large quantities of contaminating leukocytes (49).

As the comparison of established assay findings with clinical outcome will define the clinically relevant threshold, clinical assays are only useful when the results can be accurately correlated with clinical outcome of cancer patients and reflect the nature of the disease (12). In our study, we showed for the first time the prognostic significance of CTC gene expression analysis when blood has been processed 24 to 30 hours postcollection in context of NCCTG MBC clinical trials, suggesting that our findings are very promising for establishing gene expression–based CTC assays for clinical utility as prognostic/diagnostic tests in a community clinic setting. This will enable a broader application of CTC gene expression analyses/studies in clinical trials and different practice settings. As this study examined a limited number of patients with different treatment regimens and lines of therapy, our results should be interpreted as hypothesis generating.

Additional CTC gene expression analyses are ongoing in other NCCTG treatment trials including early (neoadjuvant and adjuvant settings) breast cancer, MBC and advanced lung cancer trials (50) to validate the feasibility of CTC gene expression testing in a community-based setting.

No potential conflicts of interest were disclosed.

This work was supported by NIH grants CA25224-31 and CA114740 (PI: JC Buckner) for the North Central Cancer Treatment Group and associated Biospecimen Resource, respectively, and the Breast Cancer Research Foundation (PI: E.A. Perez).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Andreopoulou
E
,
Cristofanilli
M
. 
Circulating tumor cells as prognostic marker in metastatic breast cancer
.
Expert Rev Anticancer Ther
2010
;
10
:
171
7
.
2.
Mostert
B
,
Sleijfer
S
,
Foekens
JA
,
Gratama
JW
. 
Circulating tumor cells (CTCs): detection methods and their clinical relevance in breast cancer
.
Cancer Treat Rev
2009
;
35
:
463
74
.
3.
Ring
A
,
Smith
IE
,
Dowsett
M
. 
Circulating tumour cells in breast cancer
.
Lancet Oncol
2004
;
5
:
79
88
.
4.
Hayes
DF
,
Smerage
JB
. 
Circulating tumor cells
.
Prog Mol Biol Transl Sci
2010
;
95
:
95
112
.
5.
Ignatiadis
M
,
Georgoulias
V
,
Mavroudis
D
. 
Micrometastatic disease in breast cancer: clinical implications
.
Eur J Cancer
2008
;
44
:
2726
36
.
6.
Cristofanilli
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
.
7.
Cristofanilli
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
.
8.
Hayes
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
.
9.
Riethdorf
S
,
Fritsche
H
,
Müller
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
.
10.
Budd
GT
,
Cristofanilli
M
,
Ellis
MJ
,
Stopeck
A
,
Borden
E
,
Miller
MC
, et al
Circulating tumor cells versus imaging–predicting overall survival in metastatic breast cancer
.
Clin Cancer Res
2006
;
12
:
6403
9
.
11.
Cristofanilli
M
,
Broglio
KR
,
Guarneri
V
,
Jackson
S
,
Fritsche
HA
,
Islam
R
, et al
Circulating tumor cells in metastatic breast cancer: biologic staging beyond tumor burden
.
Clin Breast Cancer
2007
;
7
:
471
9
.
12.
Riethdorf
S
,
Pantel
K
. 
Advancing personalized cancer therapy by detection and characterization of circulating carcinoma cells
.
Ann N Y Acad Sci
2010
;
1210
:
66
77
.
13.
Demel
U
,
Tilz
GP
,
Foeldes-Papp
Z
,
Gutierrez
B
,
Albert
WH
,
Bocher
O
. 
Detection of tumour cells in the peripheral blood of patients with breast cancer. Development of a new sensitive and specific immunomolecular assay
.
J Exp Clin Cancer Res
2004
;
23
:
465
8
.
14.
Hauch
S
,
Zimmermann
S
,
Lankiewicz
S
,
Zieglschmid
V
,
Bocher
O
,
Albert
WH
. 
The clinical significance of circulating tumour cells in breast cancer and colorectal cancer patients
.
Anticancer Res
2007
;
27
:
1337
41
.
15.
Reinholz
MM
,
Nibbe
A
,
Jonart
LM
,
Kitzmann
K
,
Suman
VJ
,
Ingle
JN
, et al
Evaluation of a panel of tumor markers for molecular detection of circulating cancer cells in women with suspected breast cancer
.
Clin Cancer Res
2005
;
11
:
3722
32
.
16.
Bosma
AJ
,
Weigelt
B
,
Lambrechts
AC
,
Verhagen
OJ
,
Pruntel
R
,
Hart
AA
, et al
Detection of circulating breast tumor cells by differential expression of marker genes
.
Clin Cancer Res
2002
;
8
:
1871
7
.
17.
Apostolaki
S
,
Perraki
M
,
Kallergi
G
,
Kafousi
M
,
Papadopoulos
S
,
Kotsakis
A
, et al
Detection of occult HER2 mRNA-positive tumor cells in the peripheral blood of patients with operable breast cancer: evaluation of their prognostic relevance
.
Breast Cancer Res Treat
2009
;
117
:
525
34
.
18.
Ignatiadis
M
,
Xenidis
N
,
Perraki
M
,
Apostolaki
S
,
Politaki
E
,
Kafousi
M
, et al
Different prognostic value of cytokeratin-19 mRNA positive circulating tumor cells according to estrogen receptor and HER2 status in early-stage breast cancer
.
J Clin Oncol
2007
;
25
:
5194
202
.
19.
Ignatiadis
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
.
20.
Xenidis
N
,
Perraki
M
,
Kafousi
M
,
Apostolaki
S
,
Bolonaki
I
,
Stathopoulou
A
, et al
Predictive and prognostic value of peripheral blood cytokeratin-19 mRNA-positive cells detected by real-time polymerase chain reaction in node-negative breast cancer patients
.
J Clin Oncol
2006
;
24
:
3756
62
.
21.
U.S. Department of Health and Human Services Food and Drug Administration, Center for Drug Evaluation and Research (CDER) and Center for Veterinary Medicine (CVM). Guidance for Industry
.
Bioanalytical Method Validation
2001
.
22.
Bustin
SA
,
Benes
V
,
Garson
JA
,
Hellemans
J
,
Huggett
J
,
Kubista
M
, et al
The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments
.
Clin Chem
2009
;
55
:
611
22
.
23.
Bustin
SA
. 
Absolute quantification of mRNA using real-time reverse transcription polymerase chain reaction assays
.
J Mol Endocrinol
2000
;
25
:
169
93
.
24.
Pfaffl
MW
. 
A new mathematical model for relative quantification in real-time RT-PCR
.
Nucleic Acids Res
2001
;
29
:
e45
.
25.
McShane
LM
,
Altman
DG
,
Sauerbrei
W
,
Taube
SE
,
Gion
M
,
Clark
GM
. 
REporting recommendations for tumour MARKer prognostic studies (REMARK)
.
Br J Cancer
2005
;
93
:
387
91
.
26.
Therasse
P
,
Arbuck
SG
,
Eisenhauer
EA
,
Wanders
J
,
Kaplan
RS
,
Rubinstein
L
, et al
New guidelines to evaluate the response to treatment in solid tumors. European Organization for Reseach and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada
.
J Natl Cancer Inst
2000
;
92
:
205
16
.
27.
Tewes
M
,
Aktas
B
,
Welt
A
,
Mueller
S
,
Hauch
S
,
Kimmig
R
, et al
Molecular profiling and predictive value of circulating tumor cells in patients with metastatic breast cancer: an option for monitoring response to breast cancer related therapies
.
Breast Cancer Res Treat
2009
;
115
:
581
90
.
28.
Obermayr
E
,
Sanchez-Cabo
F
,
Tea
MK
,
Singer
CF
,
Krainer
M
,
Fischer
MB
, et al
Assessment of a six gene panel for the molecular detection of circulating tumor cells in the blood of female cancer patients
.
BMC Cancer
2010
;
10
:
666
.
29.
Shen
C
,
Hu
L
,
Xia
L
,
Li
Y
. 
The detection of circulating tumor cells of breast cancer patients by using multimarker (Survivin, hTERT and hMAM) quantitative real-time PCR
.
Clin Biochem
2009
;
42
:
194
200
.
30.
Van der Auwera
I
,
Peeters
D
,
Benoy
IH
,
Elst
HJ
,
Van Laere
SJ
,
Prové
A
, et al
Circulating tumour 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
.
31.
Fehm
T
,
Braun
S
,
Muller
V
,
Janni
W
,
Gebauer
G
,
Marth
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
.
32.
Núñez-Villar
MJ
,
Martínez-Arribas
F
,
Pollán
M
,
Lucas
AR
,
Sánchez
J
,
Tejerina
A
, et al
Elevated mammaglobin (h-MAM) expression in breast cancer is associated with clinical and biological features defining a less aggressive tumour phenotype
.
Breast Cancer Res
2003
;
5
:
R65
70
.
33.
Span
PN
,
Waanders
E
,
Manders
P
,
Heuvel
JJ
,
Foekens
JA
,
Watson
MA
, et al
Mammaglobin is associated with low-grade, steroid receptor-positive breast tumors from postmenopausal patients, and has independent prognostic value for relapse-free survival time
.
J Clin Oncol
2004
;
22
:
691
8
.
34.
Pachmann
K
,
Clement
JH
,
Schneider
CP
,
Willen
B
,
Camara
O
,
Pachmann
U
, et al
Standardized quantification of circulating peripheral tumor cells from lung and breast cancer
.
Clin Chem Lab Med
2005
;
43
:
617
27
.
35.
Giribaldi
G
,
Procida
S
,
Ulliers
D
,
Mannu
F
,
Volpatto
R
,
Mandili
G
, et al
Specific detection of cytokeratin 20-positive cells in blood of colorectal and breast cancer patients by a high sensitivity real-time reverse transcriptase-polymerase chain reaction method
.
J Mol Diagn
2006
;
8
:
105
12
.
36.
Kahn
HJ
,
Presta
A
,
Yang
LY
,
Blondal
J
,
Trudeau
M
,
Lickley
L
, et al
Enumeration of circulating tumor cells in the blood of breast cancer patients after filtration enrichment: correlation with disease stage
.
Breast Cancer Res Treat
2004
;
86
:
237
47
.
37.
Taubert
H
,
Blümke
K
,
Bilkenroth
U
,
Meye
A
,
Kutz
A
,
Bartel
F
, et al
Detection of disseminated tumor cells in peripheral blood of patients with breast cancer: correlation to nodal status and occurrence of metastases
.
Gynecol Oncol
2004
;
92
:
256
61
.
38.
Lang
JE
,
Mosalpuria
K
,
Cristofanilli
M
,
Krishnamurthy
S
,
Reuben
J
,
Singh
B
, et al
HER2 status predicts the presence of circulating tumor cells in patients with operable breast cancer
.
Breast Cancer Res Treat
2009
;
113
:
501
7
.
39.
Jatana
KR
,
Balasubramanian
P
,
Lang
JC
,
Yang
L
,
Jatana
CA
,
White
E
, et al
Significance of circulating tumor cells in patients with squamous cell carcinoma of the head and neck: initial results
.
Arch Otolaryngol Head Neck Surg
2010
;
136
:
1274
9
.
40.
Lin
YC
,
Chen
SC
,
Hsueh
S
,
Lo
YF
,
Chow-Wu
YH
,
Liaw
IC
, et al
Lack of correlation between expression of human mammaglobin mRNA in peripheral blood and known prognostic factors for breast cancer patients
.
Cancer Sci
2003
;
94
:
99
102
.
41.
Bitisik
O
,
Saip
P
,
Saglam
S
,
Derin
D
,
Dalay
N
. 
Mammaglobin and maspin transcripts in blood may reflect disease progression and the effect of therapy in breast cancer
.
Genet Mol Res
2010
;
9
:
97
106
.
42.
Miller
KD
,
Chap
LI
,
Holmes
FA
,
Cobleigh
MA
,
Marcom
PK
,
Fehrenbacher
L
, et al
Randomized phase III trial of capecitabine compared with bevacizumab plus capecitabine in patients with previously treated metastatic breast cancer
.
J Clin Oncol
2005
;
23
:
792
9
.
43.
Wülfing
P
,
Borchard
J
,
Buerger
H
,
Heidl
S
,
Zänker
KS
,
Kiesel
L
, et al
HER2-positive circulating tumor cells indicate poor clinical outcome in stage I to III breast cancer patients
.
Clin Cancer Res
2006
;
12
:
1715
20
.
44.
Weigelt
B
,
Bosma
AJ
,
Hart
AA
,
Rodenhuis
S
,
van ‘t Veer
LJ
. 
Marker genes for circulating tumour cells predict survival in metastasized breast cancer patients
.
Br J Cancer
2003
;
88
:
1091
4
.
45.
Benoy
IH
,
Elst
H
,
Van Dam
P
,
Scharpé
S
,
Van Marck
E
,
Vermeulen
PB
, et al
Detection of circulating tumour cells in blood by quantitative real-time RT-PCR: effect of pre-analytical time
.
Clin Chem Lab Med
2006
;
44
:
1082
7
.
46.
Coumans
FA
,
Doggen
CJ
,
Attard
G
,
de Bono
JS
,
Terstappen
LW
All circulating EpCAM+CK+CD45- objects predict overall survival in castration-resistant prostate cancer
.
Ann Oncol
2010
;
21
:
1851
7
.
47.
Hekimian
K
,
Stein
E-L
,
Pachmann
K
,
Pachman
K
. 
Detection of tumor stem cells among circulating epithelial tumor cells
.
7th International Symposium on Minimal Residual Cancer
2009
;
76
:
abstr 012
.
48.
You
F
,
Roberts
LA
,
Kang
SP
,
Nunes
RA
,
Dias
C
,
Iglehart
JD
, et al
Low-level expression of HER2 and CK19 in normal peripheral blood mononuclear cells: relevance for detection of circulating tumor cells
.
J Hematol Oncol
2008
;
1
:
2
.
49.
Sieuwerts
AM
,
Kraan
J
,
Bolt-de Vries
J
,
van der Spoel
P
,
Mostert
B
,
Martens
JW
, et al
Molecular characterization of circulating tumor cells in large quantities of contaminating leukocytes by a multiplex real-time PCR
.
Breast Cancer Res Treat
2009
;
118
:
455
68
.
50.
Reinholz
M
,
Mandrekar
SJ
,
Foster
N
,
Meyers
JP
,
Kitzmann
KA
,
Jett
JR
, et al
Circulating tumor cell cytokeratin-19 gene expression as a prognostic factor in lung cancer: Analysis based on North Central Cancer Treatment Group (NCCTG) clinical trials (N0423/426)
.
J Thorac Oncol
2011
;
6
:
S1076
S7
.