Circulating Tumor Cells in the Management of Breast Cancer

Circulating Tumor Cells in the Management of Breast Cancer

Circulating Tumor Cells in the Management of Breast CancerGovardhanan Nagaiah,1 Jame Abraham2

Clinical Breast Cancer, Vol. 10, No. 3, 209-216, 2010; DOI: 10.3816/CBC.2010.n.028Clinical Breast Cancer June 2010There is growing evidence that circulating tumor cells (CTC) could help predict recurrence and effectiveness of therapy. The utility of CTC testing in the management of breast cancer is unclear based on current studies. Breast cancer is the most common cancer among women in the Western world after skin cancer. breast cancer carries the second highest mortality rate among woman with cancer Although some metastatic breast cancer (MBC) is considered incurable, a variety of treatment options are available that can potentially improve survival rate and quality of life. A reliable and reproducible blood test that identifies patients who will benefit from therapy would be valuable to oncologists in making treatment decisions Such a test could also be used to monitor response to treat-ment . Patients who respond may elect to continue treatment for a longer time. Conversely, patients who are not responding could be taken off toxic treatment earlier Circulating tumor cells (CTC) detection appears to have the characteristics of such a test and shows some promise. The presence of breast cancer tumor cells in the bone marrow is thought to cause consequent metastatic disease and has been shown to be a poor prognostic factor.These cells have been called disseminated tumor cells (DTC) One method of metastatic spread is the migration of tumor cells from this bone marrow reservoir and the primary site by the blood stream to loci of new metastatic lesions.These circulating tumor cells were identified as early as 1869 to be present in patients with cancerBreast cancer CTC has been defined by various markers. CTC/DTC defined by epithelial cell markers has been demons-trated to give prognostic information. Because the number of CTC in the peripheral blood is small, various methods such as antibody-coated immunomagnetic beads and membrane filters are used to concentrate these epithelial cells. CTC is detected either by morphologic methods based on cell surface antiepithelial antibodies or by nonmorphologic nuclear methods, such as reverse transcriptase polymerase chain reaction (RT-PCR).

In the membrane micro-filter assay method, after separation by a polycarbonate filter, the larger CTC are then detected by RT-PCR performed by electrolysis on the capturing membraneThe AdnaTest BreastCancerSelect (AdnaGen; Langenhagen, Germany) uses a proprietary mixture of immunomagnetic beads coated with antibodies to epithelial surface antigens to concentrate the CTC. The number of CTC is indirectly determined by a semi quantitative RT-PCR method using probes for 3 epithelial cellassociated mRNAs: mucin 1 (MUC1), HER2, and the surface glycoprotein GA 733-2.12 This test is commercially available in Europe. This mRNA-based test is thought to have higher sensitivity than whole-blood morphologybased tests such as CellSearch.

In the microfluidic CTC chip method, whole blood is passed by laminar flow into a CTC chip that has an array of microposts coated with antiepithelial cell adhesion molecule (EpCAM) antibodies.

These captured cells are stained with a nuclear marker for DNA content and an anticytokeratin epithelial cytoplasmic marker. The counting of the cells is done by fluorescence microscopyAnother recent technology is fiber optic array scanning technology.

This system uses fluorescence cytometry combined with an automated digital microscopy imaging system with laser-printing optics that scan 300,000 cells per second to detect immunofluorescently labeled CTC on a glass slide The epithelial ImmunoSpot (EPISPOT) method detects only viable CTC by testing for epithelial-associated proteins after a 48-hour cell culture. Mostert et al have described the various technologies used in CTC detection in a recent review article.

Many of these techniques, however, are cumbersome and not easily reproducible in clinical practice.Veridex, LLC, a Johnson & Johnson cancer diagnostic subsidiary, has devised a system called CellSearch to detect CTC.

CellSearch has been validated and its reproducibility demons-trated in three separate prospective clinical trials. Furthermore, CellSearch has been cleared by the US Food and Drug Administration for use in monitoring patients with metastatic breast, prostate, and colorectal cancer. The CellSearch system defines breast cancer CTC as nucleated cells that lack CD45 (to exclude leukocytes) but express cytokeratin 8, 18, or 19. It is a semiautomated immunomagnetic and immuno-fluorescent detection system. Centrifuged serum is processed through a magnetic activated cell enrichment system improving efficacy. This approach is integrated with a downstream fluorescent labeling. The analytical sensitivity is reported at 1 CTC/7.5 mL whole blood and the analytical specificity at 99.7%.Riethdorf et al did an independent study to test the robust-ness of the CellSearch system. The intra-assay and interassay precisions for different sites were high. Spiked samples were recovered between 80% and 82% of times. CTC was detected in about 70% of MBC patientsStorage of blood samples for 24, 48, or 72 hours at room temperature or at 4C or shipment did not affect results significantly.

The authors concluded that the CellSearch system was a reliable method to detect CTCBidard et al looked at the prognostic value of DTC in bone marrow and compared it with CTC. Cells were screened for pancytokeratin mAb A45-B/B3, according to the International Society of Hematotherapy and Graft Engineering classification in 138 patients. Thirty-seven of these patients also had CTC tested. A total of 59% had DTC that was associated with bone metastasis (P = .0001) but not with a poorer overall survival (OS). The subgroup of 37 metastatic patients who had CTC had a shorter survival (P = .01).Slade et al did a study in the United Kingdom to compare CTC monitoring by the CellSearch system with DTC tested by immunocytochemistry and RT-PCR for detection of micrometastasis.

Normal patients (group I, n = 210), low-risk patients (group II, T1 N0, n = 13), and high-risk patients (group III, > 3 nodes positive, n = 33) were identified and monitored for 2 years on routine follow-up with no active disease. Group IV consisted of patients with metastatic disease. All group I samples were negative for CTC. In contrast, 7 out of 18 (39%) group II primary patients and 23 out of 33 (70%) group III patients were positive for CTC (P = .042).

Patients who were consistently negative for DTC could have positive CTC, but patients who had positive DTC always had positive CTC, although these relationships were not statistically significant.

This study is limited by different methods used to detect CTC and DTC.The detection of cytokeratin 19 mRNA-positive CTC and DTC using RT-PCR was studied recently. A total of 165 patients with stage I-II breast cancer had paired samples of blood and bone marrow taken before initiation of adjuvant chemotherapy and samples were taken in 84 patients after the initiation of chemotherapy

There was a 93.9% (McNemar; P = .344) and 72.6% (McNemar; P = .999) concordance between blood and bone marrow CTC/DTC samples before and after chemotherapy, respectively. CTC and DTC were detected in 55.2% and 57.6% of the patients before chemotherapy.

After chemotherapy, CTC and DTC were identified in 44 (52.4%) and 43 (51.2%) of the 84 patients, respectively.

The detection of CTC or DTC before chemotherapy was associated with decreased OS (P = .024 and P = .015, respectively).The detection of DTC has been shown to confer higher risk of recurrence and disease-specific death in a pooled analysis of several prospective studies using various methods.

However, DTC measurement has not been accepted into clinical practice possibly because of patient reluctance to multiple bone marrow aspiration examinations.

The above study shows excellent concordance between CTC and DTC detected by RT-PCR.

Peripheral blood draws, which could be used to guide treatment options, could be more acceptable to patients to monitor their risk levels for recurrence.In a study performed at Baylor University Medical Center, Biggers et al evaluated the presence of CTC in patients with localized breast cancer using a CellSearch assay.

Patients undergoing surgery for clinically localized breast cancer were prospectively studied for CTC.

The investigators found that in a cohort of 41 patients, 10 patients (24%) had CTC preoperatively and 9 patients (30%) had CTC postoperatively, with an overall positivity for CTC of 39%.Lang et al looked at patients with CTC with operable cancers to identify relationships between CTC and tumor pathology, nodal or bone marrow spread, using the CellSearch assay.

Of the 92 patients studied, 35 patients (38%) had CTC with a median number of 1. Tumor size, estrogen receptor status, progesterone receptor status, grade, histologic type, degree of nodal involvement, lymphovascular invasion, Ki-67, and marrow micrometastases did not have a significant association with presence of CTC.

HER2 status was the only factor that reliably predicted the presence of CTCIgnatiadis et al looked into the prognostic value of the molecular detection of CTC using 3 markerscytokeratin 19, mammaglobin A, and HER2in 175 women with stage I-III breast cancer before the initiation of adjuvant chemotherapy by RT-PCR.

Univariate analysis showed that the detection of cytokeratin 19 mRNA+, mammaglobin A mRNA+, and HER2 mRNA+ cells was associated with shorter disease-free survival (DFS; P < .001, P = .001, and P < .001, respectively),

whereas the detection of cytokeratin 19 mRNA+ and mammaglobin A mRNA+ cells was associated with poorer OS (P = .044 and P = 0.034). Estrogen receptornegative tumors and the detection of cytokeratin 19 mRNA+ and mammaglobin A mRNA+ CTC were independently associated with poorer DFS in multivariate analysis.Pierga et al tested for CTC pre- and post-adjuvant therapy in patients with large operable or locally advanced breast cancer, using a group of 20 MBC patients as control for a period of 18 months.

At least one tumor cell was detected in 23% of pre neoadjuvant chemotherapy samples, 17% in post-neoadjuvant chemothe rapy samples, and 27% in all 118 patients. After a median follow-up of 18 months, the presence of CTC (P = .017), hormone receptor negativity, and large tumor size were independent prognostic factors for shorter distant metastasisfree survivalAnother study evaluated CTC at the time of primary diagnosis and during adjuvant therapy in 1767 node-positive breast cancer patients as part of the SUCCESS Trial (University of Munich, Germany). Follow-up CTC was assessed in 852 of these patients.

In this study, 10% of patients had more than one CTC before the start of systemic treatment. The presence of CTC did not correlate with breast cancer size (P = .07), grading (P = .30), hormonal status (P = .54), or HER2 status of the primary tumor (P = .26). However, CTC was found to significantly correlate with the presence of lymph node metastases (P = .02).

None of the 24 healthy control individuals showed > 1 CTC.

Among the 852 patients with follow-up CTC after the completion of treatment, 11% were CTC positive before starting systemic treatment, and 7% had > 1 CTC after completion of treatment.

Among initially CTC-positive patients, 10% remained positive and 90% had a negative CTC after chemotherapy.

Of the initially CTC-negative patients, 93% remained negative, and 7% had a positive CTC (P = .24).

Follow-up at a median of 12 months by the same authors showed 21 recurrences, and 7 patients died of their disease.

Presence of CTC before systemic treatment did not show prognostic relevance for DFS (P = .89) or OS (P = .71).

Persistence of CTC after chemotherapy was a significant predictor for both reduced DFS (P = .04) and OS (P = .03)Another German study evaluated CTC testing in the neoadjuvant setting. CTC was measured before starting systemic therapy in 245 patients with locally advanced breast cancer and after treatment in 67 patients.

At least 1 CTC was detected in 22% of the patients before starting treatment. CTC was detected in only 10.4% of the patients after treatment. CTC tests before and after systemic treatment were compared in 43 patients In this study, 10 patients (23%) had detectable CTC before starting treatment and negative CTC after finishing treatment, 6 patients (14%) developed detectable CTC post-treatment after having had no CTC before treatment, and 27 patients (63%) were CTC negative both before and after treatment.

This study shows that CTC can be detected in non-MBC patients at primary diagnosis and also after primary systemic treatment of locally advanced disease.In 91 patients with early breast cancer, relapse was highly correlated with increasing CTC after chemotherapy.

The detection of CTC is inconsistent across studies in early breast cancer.

Serial measurement of CTC might be useful when interpreted in conjunction with radiologic tests to predict response to therapy and survival.Serrano et al studied the prognostic significance of CTC detected by immunomagnetic methods in early and metastatic breast cancer patients.

Sixty-six percent of patients had CTC. Patients with > 6 CTC had a poorer prognosis. CTC was detected in 54.2% of 59 patients after 4 weeks of chemotherapy.29 Nole et al did a study to evaluate the prognostic significance of CTC detection in patients with advanced breast cancer using the CellSearch system in 80 patients.

Forty-nine patients had 5 CTC at baseline. Patients with 5 CTC at last available blood draw showed a statistically significant higher risk of progression (hazard ratio [HR], 6.4; 95% CI, 2.8-14.6) and had 5 times the risk of progression when compared with patients with 0-4 CTC at the same time point (HR, 5.3; 95% CI, 2.8-10.4).Dawood et al did a retrospective study that included 185 patients with newly diagnosed MBC who were evaluated between 2001 and 2007.

CTC were isolated and enumerated before patients started first-line treatment using the CellSearch system.

Fifty-six patients (30.3%) presented with de novo metastatic disease, and 129 patients (69.7%) presented with newly recurrent breast cancer. A total of 114 patients (61.6%) had CTC < 5, and 71 patients (38.4%) had CTC 5.

The median OS was 28.3 months and 15 months (P < .0001) for patients with CTC < 5 and CTC 5, respectively

Superior survival among patients with CTC < 5 was observed regardless of hormone receptor and HER2/neu status, site of first metastases, or whether the patient had recurrent or de novo metastatic disease.

In the multivariate model, patients with CTC 5 had a death HR of 3.64 (95% CI, 2.11-6.30) compared with patients with CTC < 5.Cristofanilli et al tested 177 patients with MBC for the number of CTC before and after a new line of treatment and compared imaging studies to evaluate disease status every 9-12 weeks.

The control group comprised healthy women and women with benign breast diseases.

The study showed that MBC patients with a baseline (before starting treatment) CTC 5 (per 7.5 mL of whole blood) had a shorter median progression-free survival (PFS; 2.7 months vs. 7.0 months; P < .001) and shorter OS (10.1 months vs. 18 months; P < .001).

CTC at baseline and at first follow-up were the most signifi-cant predictors of PFS and OS by multivariate regression analysisHayes et al did a subset analysis of the same data and demonstrated that detection of CTC at anytime during therapy predicts rapid disease progression and mortality for patients with MBC

The subset analysis was done on 83 (of the 177) patients in the same group who were receiving first-line treatment for metastatic disease.

CTC was assessed in these patients at baseline and monthly thereafter for up to 6 months, for a median follow-up of 12.2 months. Forty-three patients (52%) had 5 CTC at baseline. Patients with 5 CTC at baseline and at first follow-up (4 weeks) had a poorer prognosis than patients with 5 CTC.

CTC before and after starting therapy were strong, independent prognostic factors for both PFS and OS

In the abovementioned studies, the CTC detection rate has been around 50%.

When CTC is detected, it appears to confer a higher risk of recurrence and decreased survival.

These results have to be taken in the context that the other half of patients who did not have CTC did progress eventually, albeit more slowly; therefore, we could be missing other factors that contribute to relapse.A series of 177 patients with MBC were evaluated for correlation between MUC1 protein (carcinoma antigen [CA] 27.29 and/or CA 15-3) and CTC in 65 patients who underwent both tests at serial time points.

The study showed that 37% of the patients had 5 CTC, and 29% had a > 25% increase in MUC1 at first follow-up.

CTC and change in MUC1 did not significantly correlate at any time.

CTC correlated with OS (P = .006), and both MUC1 and CTC correlated with PFSIn a community-based practice, 50 patients with MBC were evaluated for both CTC and CA 27.29 at baseline and monthly for 6 months.

CTC was detected in 54% of the patients with MBC, compared with 70% in the Cristofanilli et al study.

In this study CTC testing demonstrated high specificity (89%) and less sensitivity (70%) to detect disease progression shown by radiologic tests.

Conversely the traditional biomarker CA 27.29 had a higher sensitivity of 85% but was less specific (31%). Patients with < 5 CTC 1 month after starting treatment had a longer median PFS (P = .023).

The CA 27.29 level failed to show a similar difference.

The authors of this study suggested that in contrast to CA 27.29, high levels of CTC could be reflecting tumor biology rather than tumor burden.A small retrospective analysis showed a statistically significant correlation among computed tomography (CT)positron emission tomography (PET) scans, tumor marker CA 27.29, and CTC test numbers.

However, they still found an unexplained discordance of these test results.

That is, the same result of the CT-PET scan was occasionally associated with various results of both the CA 27.29 test and CTC testing in different patients.

This discordance could again be reflecting the diverse tumor biology of MBCSerial CTC by CellSearch were collected in MBC patients at the start of new treatments including chemotherapy, endocrine, and biologic therapy.

Disease progression occurred in 29% of cases with < 5 CTC per 7.5 mL of blood and 69% of cases with > 5 CTC per 7.5 mL of blood (P < .0001).

When radiologic and CTC assessments were correlated from the same patient, > 5 CTC per 7.5 mL of blood is associated with a 5.32-fold higher risk of disease progression (P < .0001) In addition, > 5 CTC per 7.5 mL of blood is associated with higher sensitivity and specificity than lactate dehydrogenase, alkaline phosphatase, and the MUC1 glycoprotein in predicting disease progression in patients with measurable MBC

A total of 151 patients with MBC had CTC tested. The prognostic significance of CTC was compared with Swenerton score, CA 27.29 level, age (< 50 years vs. 50 years), hormone receptor status, and HER2 status, metastatic site, and type and line of therapy. It was found that 44% of patients had > 5 CTC with median OS of 13.5 months.

Patients with < 5 CTC survived for 29.3 months (P < .0001).

In patients with 5 CTC the HR of death was 2.2 (P = .003)The prognostic relevance of fluorine-18 [18F] fluorodeoxyglucose (FDG) CT-PET scan was compared with CTC in a recently published study of a retrospective analysis of 115 MBC patients.

Both tests were done at baseline and at 9 and 12 weeks of therapy (midtherapy).

In 102 evaluable patients, midtherapy CTC correlated with FDGCT-PET response in 68 patients (67%).

Midtherapy CTC counts and FDGCT-PET response predicted OS (P = .001)In a total of 9 patients followed for 9.8 months, the sensitivity of CTC to detect metastatic disease shown on CT-PET was 32% and specificity was 100%.

In the above studies, CTC appears to correlate with radiologic tests but is not sensitive enough to replace radiologic testing.

The relationship between CTC and breast cancer tumor makers is unclear and needs further investigation.DiscussionCirculating tumor cells have been detected in all stages of breast cancer. In general, the presence of 5 CTC appears to confer a high risk at any stage.

The detection rate of CTC in early breast cancer is not consistent across studies. This could be because of the various techniques used to detect CTC. Even though the CellSearch method appears to be more commonly used, RT-PCR might be more sensitive.

In a recently reported study, CTC was detected in 55.2% of the patients with stage I and II breast cancer by RT-PCR for cytokeratin 17.

CellSearch has the advantage of being semiautomated and consistently reproducible.

When used in clinical practice, the method of CTC detection has to be chosen after careful thought depending upon locally available resources.Circulating tumor cells could be used as additional tests to risk stratify patients with early breast cancer.

CTC are thought to represent chemo-resistant cells that potentially could contribute to recurrence later.

Therefore, in patients with early breast cancer, serial measurements of CTC could help identify patients in whom additional therapy might have to be considered.

For example, because there is concordance between positive CTC and radiologic testing, CTC could be done more often, and a trend of increasing CTC could prompt aggressive radiologic follow-up. CTC could be tested in MBC patients at the initiation of new therapy.

CTC then could be tested after 1 or 2 cycles.

Change in CTC appears to precede radiologic changes that are known to occur only after several weeks or months of therapy.

CTC < 5 or a decreasing or stable trend in CTC could represent effectiveness of therapy. On the other hand, an increasing trend of CTC could represent early progression and prompt a change in therapy.

It would be prudent to wait at least two cycles before changing therapy.

Because hormonal therapy works more slowly than chemotherapy the optimal time for CTC evaluation could be 6-8 weeks after start of therapy.

Patients on chemotherapy could have CTC tested 3-5 weeks after initiation of treatment. CTC levels that are consistently < 5 could represent good outcomes particularly when they are associated with improvement or stability in radiologic testing.

Another interesting role of CTC testing could be in determining HER2 status.

Recent studies showed discordance between HER2 status of the primary tumor and CTC.In 66 patients with advanced breast cancer, 29% (8/28) of HER2-negative primary tumors had HER2-positive CTC and 42% (5/12) of HER2-positive primary tumors had HER2-negative CTC

In this setting, CTC could help in determining therapy.

CTC measurements should also be interpreted with caution, as not all breast cancer CTC might be detected by the presently available detection methodology. A recent study reported that around 10% of patients have a normal genotype of invasive breast cancer that is typically negative for EpCAM expression.

CTC from these tumors could be potentially missed by CTC technologies such as CellSearch that use EpCAM to concentrate and detect CTC. An ongoing trial conducted by the Southwest Oncology Group and the Breast Cancer Intergroup of North America involving patients with stage IV breast cancer is designed to address the specific question of whether changing patients to an alternative chemotherapy regimen after 3 weeks of first-line chemotherapy on the basis of an elevated CTC (> 5 per 7.5 mL of blood) affects the PFS in patients with MBC The planned sample size of 500 patients will be divided into three groups.

All patients will have CTC taken at baseline before their first course of chemotherapy.

Group 1 patients with < 5 CTC at baseline will receive regular treatment without change at the discretion of the treating physician and no further blood will be collected for CTC. Patients with > 5 CTC at baseline will undergo a second blood draw at day 22 after the completion of one course of chemotherapy.

Patients with < 5 CTC will be assigned to group 2 and will receive their current chemotherapy with no change.

Patients with > 5 CTC after their first line chemotherapy will be randomized into either continuing their current therapy or switching to a different chemotherapy as decided by the treating physicianConclusionThe American Society of Clinical Oncology guidelines released in 2007 do not recommend CTC or DTC testing methods to risk stratify breast cancer patients.

There are multiple clinical trials at various stages of accrual with results being awaited.

There are multiple competing technologies with varying sensitivities and specificities that are being used in the detection of CTC at the present time complicating the interpretation of the results. Most of the CellSearch studies have used > 5 CTC as an arbitrary cut off for a positive result.

Whether greater number of CTC confers even greater risk of relapse has not been rigorously studied.

Even with this cutoff of > 5 CTC, it is unclear if > 5 CTC by CellSearch has the same clinical implication as the same result by a different assay. Internet search reveals multiple companies offering CTC testing to oncologists.

There have been no studies on whether CTC results obtained outside the investigational setting have any prognostic value.

There is a risk of overtreatment or undertreatment in providing clinicians with a positive test result before these methodological issues are resolved.

The testing of additional markers associated with CTC such as HER2 is being investigated. It is therefore prudent to wait for more data before recommending CTC as a risk predictor in clinical practice. Until there are prospective data available from ongoing studies CTC testing should be considered investigational.

In the future, CTC testing in combination with tumor markers, radiologic testing, and gene profiling could help clinicians determine with greater precision the usefulness of specific therapies for patients with breast cancer.

However, at the present time testing and interpretation of CTC continues to be an evolving science and should be approached with caution.