Chasing CTCs: Novel Technologies May Unlock Potential of Elusive Biomarker

Daniel F. Hayes, MD

Despite being discovered more than 150 years ago, tumor cells present in the blood of patients with cancer are only now inspiring significant research efforts. Technological advancements have allowed the isolation and enrichment of these rare cells and, as potential metastatic “emissaries,” they have significant potential for improving the detection and treatment of advanced and possibly even early-stage disease.

While a plethora of studies have demonstrated prognostic value in the number of circulating tumor cells (CTCs) present in a patient’s bloodstream, the ability to use CTCs to inform clinical decision-making remains elusive. Newer analytical techniques and single-cell molecular profiling are contributing to an improved understanding of CTCs that is already being applied in clinical trials. Leading experts propose that while we aren’t quite there yet, promising data are set to mature in the coming years.

Evolution of CTC Technology

In the mid-1800s, an Australian physician noted “cells identical with those of the cancer” in the blood of a patient with metastatic disease. Scientists at the time postulated that these CTCs, as they came to be known, had broken off of the original tumor and might supply the “seeds” for the metastatic spread of cancer by traveling through the bloodstream and lymphatic system and lodging in capillaries around distant organs.

Further study of CTCs was hindered by the fact that they are so rare—there are typically 1 to 10 CTCs per 10 mL of blood—so it was difficult to isolate them from the mass of other cells in a blood sample. It wasn’t until relatively recently that effective methods were developed to routinely isolate and enrich CTCs from blood samples of patients with cancer and CTCs began to receive wider attention.

Thanks in large part to the development of a wide range of companion analytical technologies for CTCs since the 1980s, there has been an explosion in the number of CTC-related scientific publications. According to Daniel F. Hayes, MD, who has been involved in CTC work for 15 years, at last count there were more than 40 CTC platforms at various stages of development. “Of those, there are only three or four that I think have actually been shown to have analytical validity, that if you run the assay five times you get the same results,” said Hayes, who is the Stuart B. Padnos Professor of Breast Cancer Research at the University of Michigan Comprehensive Cancer Center in Ann Arbor.

Hayes has been involved in the development of CellSearch, a Janssen Diagnostics system that is the only CTC tool that the FDA has approved. He has received research funding from CellSearch manufacturers and has current and pending patents with Janssen for novel CTC technologies.

In order to isolate and enrich CTCs, the majority of analytical technologies use a method known as positive selection, which exploits biological (eg, presence of tumor-associated antigens) or physical (eg, variations in size or density) properties that are specific to CTCs and absent in normal blood cells.

The CellSearch system, for example, takes advantage of the fact that CTCs typically express the epithelial cell—specific antigen—the epithelial cell adhesion molecule (EpCAM). It uses magnetic particles coated with antibodies that bind to Ep- CAM-expressing cells. The cells are then stained with cytokeratin and CD45 antibodies, markers of epithelial and white blood cells, respectively, and a DNA dye. The stained cells are visualized in a chamber and a computer imaging algorithm is used to identify cells that are intact, have a nucleus, and express cytokeratins, but do not express CD45. These cells are classed as CTCs. In recent years, interest has shifted toward the development of CTC technologies that employ socalled negative depletion.

Mehmet Toner, PhD, said in an interview that positive selection has several limitations that drove this shift. “You need to know something on the surface of CTCs that is absent on blood cells so you can positively capture these rare cells, but what we’ve learned over the past 10 years is that these cells are very dynamic in their phenotype,” said Toner, a professor of Surgery (Biomedical Engineering) at the Massachusetts General Hospital (MGH), Harvard Medical School, and founding director of the National Institutes of Health BioMicroElectroMechanical Systems (BioMEMS) Resource Center. “You also need to know something about the cancer behind the CTCs, which limits the utility of the assay in a broader sense.” Negative depletion assays, on the other hand, filter away red blood cells, white blood cells, and platelets, and leave only CTCs behind. “Instead of going after the needle in the haystack, as with positive selection, you get rid of the haystack and leave the needle behind,” Toner explained.

He and others at MGH developed CTC-iChip, a CTC assay based on microfluidic technology. It actually uses two separate microfluidic devices that put the blood sample through three stages combining magnetic labeling and microfluidic sorting. In this case, however, it’s the white blood cells, not the CTCs, that are magnetically labeled, so that the CTCs that are left behind have not been manipulated in any way or subjected to the potential bias of positive selection based on the expression of specific markers.

“CTC-iChip is an extremely sensitive way of sorting cells at the speed of about 20 million cells per second so that you can quickly get rid of a massive number of cells without losing the target cells,” Toner said.

This assay has successfully identified CTCs from a variety of different cancer types, including several that lost or never had epithelial markers and therefore would not have been detected by antibody- based sorting methods. According to Toner, the assay is being scaled up into randomized trials for clinical indications and they are working with a number of manufacturers, but clinical use is still a way off. Janssen Diagnostics, which entered into a collaborative agreement with MGH to establish a center of excellence for CTC research in 2011, is among the supporters of the research. Earlier iterations of CTC-iChip, called CTC-Chip and Herringbone-Chip, which use positive selection, were recently used in a study at MGH that identified clusters of CTCs in the circulation.

The clinical significance of these clusters remains unclear, but this study did determine that they were most likely made up of cells that broke off from the primary tumor in a clump and that around 2% to 5% of CTCs are likely to be found in clusters. Most significantly, they showed that CTC clusters had a significantly greater metastatic potential than single CTCs and were longer-lived in the circulation.

Current Clinical Applications of CTC Assays

As they have gained the ability to study CTCs in more detail, researchers began to recognize significant potential for a number of different clinical uses (Table 1). Numerous clinical studies have shown that simple enumeration of CTCs within a blood sample drawn from a patient with cancer can predict prognosis and this use has been the most extensively studied clinical role.

Table 1. Potential Clinical Uses of CTCs

Early Diagnosis

Identify the potential for metastasis before metastatic tumors occur or identify primary or metastatic tumors at an early stage, before they become detectable by other methods

Tumor Staging

Form part of tumor staging criteria on the basis of earlier diagnosis of metastatic disease or metastatic potential

Prognosis

Predict the likely course of disease and patient outcomes

Monitoring Response to Treatment

Determine sensitivity or resistance to a particular therapy

Guiding Treatment Decisions

Change management strategy based on the above determination and improve patient outcomes

Surrogate Endpoint

Act as a biomarker of response in clinical trials that could substitute for clinical endpoints

Personalized Treatment

Serve as a real-time “liquid biopsy” of the phenotypic and genotypic status of an individual patient’s tumor to guide treatment decisions

CellSearch is approved for this use in patients with metastatic breast cancer (MBC), colorectal cancer (CRC), and prostate cancer, predominantly on the basis of three key clinical trials in which it was used to enumerate CTCs at baseline and after therapy in patients who were about to begin chemotherapy. Across all three tumor types, it was shown that the presence of CTCs above a certain threshold (≥5 CTCs/7.5 mL blood for prostate and breast cancer and ≥3 CTCs/7.5 mL blood for CRC) were independent and accurate predictors of poorer survival (Table 2).

Some researchers have argued against the use of threshold values such as these, citing evidence that the nature of the association between CTC number and survival is continuous, at least in prostate cancer. Regardless, Hayes said Cell- Search “clearly identifies a group of patients who have a worse prognosis if they have CTCs compared to not.” This has been bolstered by numerous subsequent reports, including a recent large, prospective trial in MBC, the phase III Southwest Oncology Group (SWOG) S0500 trial, published in the Journal of Clinical Oncology in June.

Studies using CellSearch have also shown that changes in CTC number during treatment are also prognostic. Survival was improved among patients whose CTC counts decreased below the threshold level during treatment, while the converse was true for patients whose CTC counts increased during treatment. Thus, CellSearch can also be used to monitor changes in prognosis over the course of treatment in conjunction with other clinical information.

In this way, CTC levels act as a sort of surrogate biomarker of response to chemotherapy. Jeffrey B. Smerage, MD, PhD, clinical associate professor of Medical Oncology at the University of Michigan, explained, “If you follow a patient over time, typically if you were to get a baseline before they start therapy and then with each cycle, those CTCs go down when patients are responding, and when they start to come back up again it suggests that in the very near future that their cancer is going to start to progress again.”

Table 2. Key Studies Correlating Outcomes With CTC Thresholds^a

Tumor Type

Key Findings (CTCs/7.5 mL blood)

Metastatic breast cancer¹

Elevated and increasing levels of CTCs at baseline and after therapy were independent prognostic indicators of PFS and OS

• Pretreatment: Median PFS 6.5 months (≥5 CTCs) vs 11.4 months (<5 CTCs) (P <.0001); median OS 15.5 months vs 37.1 months (P <.0001)
• Posttreatment (5 weeks): Median PFS 4.8-6.7 months (≥5 CTCs) vs 9.6-10.7 months (<5 CTCs) (P <.0001); median OS 13.1-22.4 months vs 27.0-41.5 months (P <.0001)

mCRPC²

Levels of CTCs were the most accurate and independent predictor of OS^b

• Pretreatment: Median OS 11.5 months (≥5 CTCs) vs 21.7 months (<5 CTCs) (P <.0001)
• Posttreatment: Median OS 6.7-9.5 months vs 19.6-20.7 months (P <.0001)

mCRC³

Number of CTCs before and during treatment is an independent predictor of PFS and OS^b

• Pretreatment: Median PFS 4.5 months (≥3 CTCs) vs 7.9 months (<3 CTCs) (P = .0002); median OS 9.4 months vs 18.5 months (P <.0001)
• Posttreatment (range of intervals): Median PFS 1.2-3.8 months vs 6.3-7.3 months (P <.0001); median OS 3.3-6.1 months vs 14.6-15.7 months (P <.0001)

^aCellSearch system used to measure CTCs. ^bPivotal clinical data for CellSearch indications. CTCs indicates circulating tumor cells; mCRC, metastatic colorectal cancer; mCRPC, metastatic castration-resistant prostate cancer; OS, overall survival; PFS, progression-free survival.

1. Bidard FC et al. Lancet Oncol. 2014;15(4):406-414;

2. de Bono JS et al. Clin Cancer Res. 2008;14(19):6302-6309;

3. Cohen SJ et al. J Clin Oncol. 2008;26(19):3213-3221.

Hayes highlighted the potential benefit of this application of CTCs, which he said helps him to avoid repeating invasive scans that may not be necessary.

“If a patient doesn’t have symptoms or physical examination findings of progression, and if their circulating tumor markers CA15-3 and CEA are not rising, and if their CTCs are not rising, I quit doing scans.,” he said. “If any of those things look like something is going on, then I get the scans for cause as opposed to just doing them routinely.”

Both Hayes and Smerage pointed out that not everybody agrees with using CTCs in that manner, although it is an approved use of Cell- Search, and it is unclear how many clinicians actively use it that way today.

CellSearch also has been evaluated and continues to be evaluated in a number of other tumor types, including lung cancer, melanoma, and ovarian cancer. Thus far it has not demonstrated consistent analytical validity in these tumor types and thus has yet to be approved by the FDA in those indications.

Seeking Clinical Utility

Hayes is a clinical researcher with expertise in tumor biomarkers who has been closely involved with the development of tumor marker guidelines for more than a decade. He said effective biomarkers should demonstrate analytical validity (be accurate, reproducible, and available), clinical validity (distinguish between patients in the clinic), and clinical utility (direct patient care).

Enumeration of CTCs has been shown to have analytical and clinical validity, at least in the context of using CellSearch in breast, prostate, and colorectal cancers, but demonstration of clinical utility remains elusive.

Several studies are focusing on this question and, in fact, the SWOG S0500 study in MBC was actually designed for this purpose. CellSearch was used to monitor CTC levels before and after administration of first-line chemotherapy. Patients who had persistently elevated CTCs despite treatment were randomized to switch to a different chemotherapy compared with remaining on the same one.

The hope was that persistently high CTCs would indicate a failure to respond to one type of chemotherapy and therefore switching to an alternative chemotherapy would improve patient outcomes (in this case in terms of overall survival), thus demonstrating the clinical utility of CTCs. However, “It turned out that switching to a different kind of cytotoxic chemotherapy was not beneficial; it didn’t make them do any better by switching early,” said Smerage, the lead author of the study.

Though disappointing, both Hayes and Smerage speculated that these results might simply indicate that cytotoxic chemotherapy is not the ideal treatment for this group of patients.

“We should just quit kidding ourselves that giving those patients serial chemotherapy is going to help them. It’s not that I wouldn’t treat them but I think what we need to do is think about some of the new genomics tests that are available,” said Hayes. Smerage concurred. “I think CTCs may help to identify those patients where the risk of just using standard chemotherapy is relatively high and that’s a population that would benefit from novel approaches to cancer therapy,” he said.

At the same time, Smerage believes that the evidence on the clinical utility of CTCs has not yet been established. “I think in terms of being able to do a test to say what treatment you should get, circulating tumor cells aren’t quite there yet. That is going to require the data to mature from ongoing studies,” said Smerage.

Among these ongoing studies is the TREAT-CTC trial in which patients with HER2-negative primary breast cancer with detectable CTCs after neoadjuvant chemotherapy are being randomly assigned to trastuzumab or placebo (NCT01548677).

While he remains optimistic about the potential clinical utility of CTCs, Smerage did point out that the results of the SWOG 0500 trial might have an impact upon insurance coverage for testing. “I will say that recently insurance coverage has become a little more difficult,” he said.

Making the Genomic Connection

Researchers are now looking beyond enumeration to the potential clinical applications of characterizing and culturing CTCs, which has been facilitated by the development of technologies like CTC-iChip that allow isolation of viable, unmanipulated CTCs. This idea has been dubbed a “liquid biopsy” as researchers are performing the same kinds of phenotypic and genotypic analyses on these cells that they would on cells taken from a solid tumor biopsy.

Because tumors are continually evolving, acquiring new genomic alterations that impact cancer progression and treatment efficacy, serial biopsies are needed to assess the current mutational status of the tumor. In many cases, particularly in sites like the bone or the lung, there are serious risks that make it difficult to justify repeat biopsies. If CTCs reflect the current state of the tumor, as many suspect they do, then serial biopsies would simply be a case of repeat blood draws and clinicians could noninvasively analyze treatment sensitivity and resistance in real time.

Genomic analyses of CTCs are still in their infancy and, while studies have shown that CTCs express common tumor markers such as epidermal growth factor receptor and human epidermal growth factor receptor 2 (HER2), these analyses have raised as many questions as they’ve answered.

For example, several studies have demonstrated that patients with HER2-negative primary tumors may have CTCs that are HER2-positive. Clinical trials are under way to further understand these nuances. The DETECT studies are the first phase III clinical trials in which treatment decisions are being based on the phenotypic characteristics of CTCs.

DETECT III (NCT01619111) is comparing lapatinib in combination with standard therapy versus standard therapy alone in patients with HER2-negative MBC who have HER2-positive CTCs. Meanwhile, DETECT IV (NCT02035813) is enrolling patients with HER2-negative, hormone receptor—positive MBC and persistent HER2-negative CTCs. Patients will receive standard endocrine therapy and the mTOR inhibitor everolimus.

Through such studies, researchers hope to advance not only the understanding of CTCs but also how they can be used to individualize patient care.

Key Resources

Aceto N, Bardia A, Miyamoto DT, et al. Circulating tumor cell clusters are oligoclonal precursors of breast cancer metastasis. Cell. 2014;158(5):1110-1122.

Alix-Panabières C, Pantel K. Challenges in circulating tumour cell research [published online July 31, 2014]. Nat Rev Cancer. 2014;14(9):623-631.

Becker TM, Caixeiro NJ, Lim SH, et al. New frontiers in circulating tumor cell analysis: a reference guide for biomolecular profiling toward translational clinical use [published online October 29, 2013]. Int J Cancer 2014;134(11):2523-2533.

de Wit S, van Dalum G, Terstappen LW. Detection of circulating tumor cells [published online July 15, 2014]. Scientifica (Cairo). 2014;2014:810362.

Hong B, Zu Y. Detecting circulating tumor cells: current challenges and new trends. Theranostics. 2013;3(6):377-394.

Kim K, Lee KH, Lee J, Choi J. Overview of current standpoints in profiling of circulating tumor cells [published online November 9, 2013]. Arch Pharm Res. 2014;37(1):88-95.

King JD, Casavant BP, Lang JM. Rapid translation of circulating tumor cell biomarkers into clinical practice: technology development, clinical needs and regulatory requirements [published online November 5, 2013]. Lab Chip. 2014;14(1):24-31.

Liberko M, Kolostova K, Bobek V. Essentials of circulating tumor cells for clinical research and practice [published online July 5, 2013]. Crit Rev Oncol Hematol. 2013;88(2):338-356.

Parkinson DR, Dracopoli N, Petty BG, et al. Considerations in the development of circulating tumor cell technology for clinical use. J Transl Med. 2012;10:138. doi:10.1186/1479-5876-10-138. Plaks V, Koopman CD, Werb Z. Cancer. Circulating tumor cells. Science. 2013;341(6151):1186-1188.

Smerage JB, Barlow WE, Hortobagyi GN, et al. Circulating tumor cells and response to chemotherapy in metastatic breast cancer: SWOG S0500 [published online June 2, 2014]. J Clin Oncol. pii: JCO.2014.56.2561.