August 1, 2012 (Vol. 32, No. 14)
Josh P. Roberts
Circulating tumor cells (CTCs) are, by definition, epithelial cells with an intact nucleus that express cytokeratin and don’t express the leukocyte-common antigen CD45, which are found in the blood of cancer patients.
Their presence and abundance in patients who have undergone treatment has been shown to correlate with disease progression and overall survival. They are extremely rare even in such patients: estimates are in the range of 10-8–10-9, and 50 from a milliliter of blood is considered a good yield.
Yet no one truly knows what CTCs are, where they come from, what they mean, what the best way is to get at them, or even what to with them afterward. There is a single FDA-sanctioned technology, Veridex’ CellSearch, approved only for the semi-automated enumeration of CTCs from 7.5 mL of whole blood.
That doesn’t stop other techniques—some of which build upon the CellSearch platform, others utilizing distinct technologies—from generating a lot of excitement as well. Academic scientists, physician-researchers, and industry players gathered at Select Biosciences’ inaugural “Circulating Tumor Cells” conference to discuss how to capture, analyze, and interpret the data from these putative biomarkers that many feel have the potential to significantly impact clinical oncology.
Minetta Liu, M.D., associate professor at Georgetown University Medical Center, uses CellSearch for “all of my patients within the indication of metastatic disease, and I have found it useful in timing imaging studies, in helping to determine in conjunction with routine clinical factors and our radiological studies whether current therapy is helping a patient or not,” she said. “We don’t know what we’re counting, but I do know that counting the cells with that technology does correlate with patient outcomes.”
The hope is to go beyond simply enumerating CTCs. “We need to figure out what these cells are so we can manipulate them, to do better in those outcomes,” she added. Dr. Liu is involved in a clinical trial comparing molecular features of CTCs with those from primary breast tumor or metastatic sites, “trying to get an understanding of what these cells are trying to tell us.”
Stefanie Jeffrey, M.D., chief of surgical oncology research at Stanford University School of Medicine, is interested in how CTCs may be used to individualize therapy. Even from the same blood draw you can see different types of CTCs, and this may have implications for treatment, especially in a metastatic setting. One drug may be needed to treat one kind of metastasis and another drug to treat another, but you usually cannot biopsy all metastases a patient has.
“If you could do a liquid biopsy in real time, and could see how the tumor cells are changing, then potentially you may be able to more intelligently select the therapy—but this would have to be proven within a prospective clinical trial,” Dr. Jeffrey said. “We’re hoping that CTCs that are released to the bloodstream may be a surrogate, representing some of the cells that are involved in the seeding and reseeding of metastases.”
Lihua Wang, M.D., Ph.D., senior scientist of laboratory of toxicology and pharmacology at Frederick National Lab, develops CTC assays for targets that have been demonstrated for tumors, to have a way to measure the effect on tumor cells multiple times after administration of a drug—the assumption being that what’s happening in the CTCs is accurately reporting what’s happening in the tumor. But this needs to be proven. In fact, one of Dr. Wang’s fundamental goals is to establish whether a CTC is an equivalent to a biopsy.
A true surrogate not only accurately reports the same effect on its target, but reports it quantitatively as well. For the assays they have looked at there is concordance, she noted. “We have not yet proven that it’s a surrogate. That’s a high hurdle.”
Cells can be captured part way through the automated steps of CellSearch for the purpose of characterization, and there are also other ways to collect CTCs as well. These may not capture the same set of CTCs, and “it’s important to look at these side by side and to try to really figure out what they are and where they might be most useful,” remarked Dr. Liu.
Among these is the MagSweeper, an automated device that uses a reiterative process of magnetic capture, wash, and release to enrich CTCs by up to 108, yielding a highly purified sample of live CTCs that can be analyzed on a single-cell basis. The MagSweeper was invented by Dr. Jeffrey and Stanford colleagues from the Genome Technology Center and School of Engineering, and has been licensed to Illumina.
Dr. Jeffrey donates her royalties to a nonprofit charity “so that I can use other technologies that may isolate more CTCs but maybe not as pure for different purposes, or to capture different phenotypes of CTCs,” she pointed out. “And so we’re testing other technologies in our lab as well.”
The On-Q-ity platform combines affinity capture with size filtration capture to isolate CTCs. About 100,000 micropillars, conjugated with antibody, line the chamber of the plastic microfluidic device. The micropillars are arrayed in a gradient—they “get closer and closer together as the blood flows through the device, allowing you to capture the cells based on size in addition to affinity,” explained Kam Sprott, Ph.D., the company’s director of product development. The smaller leukocytes should be able to flow through even the smallest gaps, while CTCs are retained in the chamber for imaging or further processing.
Another way to achieve a similar end is to “make sure that cancer cells collide with the walls of the chip very often, but the contaminating cells—the things we don’t want to capture, which is mostly white blood cells—don’t collide with the surface anywhere near as much,” said Brian Kirby, Ph.D., who directs the Micro/Nanofluidics Laboratory in Cornell University’s Sibley School of Mechanical and Aerospace Engineering. He used his training in fluid mechanics to design a three-dimensional geometrically enhanced differential immunocapture (GEDI) microfluidic device in close collaboration with Weill Cornell Medical College’s David Nanus, M.D., and Evi Giannakakou, Ph.D.
CellSearch gives a count of EpCAM+, DAPI+, cytokeratin+, CD-45– cells, and Veridex sells research use only kits allowing the magnetic/fluorescence-based CellSearch system to identify which of these is, for example, Her-2/neu+. Researchers are also developing a host of other assays to be used with this or the more than 50 (by Dr. Liu’s count) various technologies to collect CTCs.
Some are based on staining with antibodies to intracellular and cell-surface targets before the cells are sorted. For example, Dr. Wang looks for the effect of third-generation topoisomerase I inhibitors by staining cells for phosphorylated H2AX (γH2AX), a marker of DNA double-strand breaks.
Some, like Dr. Kirby, stain the cells right on the chip after they have been captured. To demonstrate that the device had captured CTCs, for example, cells from patients with metastatic prostate cancer were fixed and stained for prostate-specific antigen, CD45, EpCAM, and DAPI and examined by confocal microscopy.
Researchers have extracted nucleic acids from CTCs. Some lyse all the cells right on the chip, using the lysate for pooled analysis expression patterns. Others, like Dr. Jeffrey’s group, for example, used a microfluidic approach to measure the expression of 87 cancer-associated and reference genes in over 100 individual CTCs isolated from breast cancer patients. Their profiles differed from one another, as well as from those of single cells taken from breast cancer cell lines—leading them to conclude that individual CTCs may be the way forward in drug discovery.
Groups may use antibodies other than an anti-EpCAM to capture different subsets of CTCs. Dr. Kirby, for example, coats his GEDI device with a monoclonal antibody recognizing the extracellular domain of prostate-specific membrane antigen (PSMA). And On-Q-ity has “developed some markers that we feel confident will recognize cells undergoing EMT,” the epithelial–mesenchymal transition thought to be involved in metastasis, said Dr. Sprott. “It may be picking up more cells in each patient, or potentially picking up patients that don’t have epithelial cells in their blood but have cells undergoing EMT.”
And various techniques can also be used to look at other cells and cell fragments that are irregular in shape, that may not be counted by CellSearch-based methods.
In addition to examining fixed or lysed cells, considerable effort is being put into vital assays and growing CTCs as well.
Dr. Kirby’s group has performed functional assays right on-chip, incubating GEDI-captured cells with taxanes for 24 hours prior to fixing and staining. They were able to demonstrate microtubule bundling in the drug-treated, compared to control, cells.
Dr. Jeffrey is trying to grow cells in vitro for drug sensitivity testing and also developing a chip to test chemoresponsiveness. “The question is whether we will be able to obtain enough CTCs, because to do drug-sensitivity testing you need to start with a certain number and measure dose response curves,” she noted. “Since the CTCs are obtained fresh, hopefully the cells will still retain features that reflect microenvironment interactions they’ve had that predispose to further cancer spread.”
Leveraging Mass Spec
Experiments simultaneously querying up to 100 intracellular and cell surface markers on 1,000s of cells per minute are within reach, said Scott Tanner, Ph.D., co-founder and CTO of DVS Sciences, who spoke at the recent “International Society for Advancement of Cytometry” conference.
The company’s CyTOF® Mass Cytometer “looks and feels like a flow cytometer, but happens to have an exquisitely sensitive atomic mass spectrometry engine,” he explained. Data is even stored in standard FCS file format, allowing it to be analyzed by standard software.
Instead of staining cells with fluorophore-conjugated antibodies, though, antibodies are attached to polymers with rare elemental metals—typically about 120 atoms per antibody. Samples entering the device are vaporized, atomized, and ionized as they flow through 7,000 degree argon plasma, and the ions are read by mass spec. “I simply count the number of atoms of each stable isotope that I used as a tag in each cell.”
Because the signals from the isotopes don’t overlap, “there’s one titration for as many parameters as you want to run at the same time, and we don’t have to make them all the same intensity,” Dr. Tanner said. “You don’t have to worry about compensation.”