Identifying Metastatic Breast Cells from Peripheral Blood

Institution: Lawrence Livermore National Laboratory
Investigator(s): Kristen Kulp, Ph.D. -
Award Cycle: 2005 (Cycle 11) Grant #: 11NB-0178 Award: $490,774
Award Type: IDEA Competitive Renewal
Research Priorities
Biology of the Breast Cell>Pathogenesis: understanding the disease



Initial Award Abstract (2005)
The early and accurate detection of metastatic breast disease would constitute a significant advance in the treatment of breast cancer. Currently, the presence of cancer in the regional lymph nodes remains the most valuable prognostic indicator for metastasis; however, as many as 30% of patients considered to be node negative eventually develop recurrent disease. Because breast cancers have been shown to shed tumor cells into the circulation soon after they become invasive, it has been proposed that detection of circulating tumor cells (CTCs) in the blood provides an index of the risk for developing distant metastases.

The hypothesis of this project is that metastatic breast tumor cells circulating in the peripheral blood of breast cancer patients can be positively identified by statistical analysis of “molecule-specific images” of individual cells. We are using Time-of-Flight Secondary Ion Mass Spectrometry (ToF-SIMS) to measure changes in cells that are characteristic of metastases. Metastatic progression results from altered genes, which produce altered protein products that can be measured as changes to the mass spectrum of the cell components. To provide the groundwork for later translational studies in humans, we are exploring the feasibility, sensitivity, and refinement of our approach using animal (rat) models of breast cancer.

In the first CBCRP grant, we showed that differences in MTLn3 (highly metastatic) and MTC (non-metastatic) cells could be detected using principal component analysis (PCA) of ToF-SIMS spectra from single cells. These breast cancer cells were originally obtained from the same tumor, so they have a similar genetic background, but they differ in their ability to form metastases. These data demonstrate that it is possible to differentiate cells by their subtle chemical differences in disease potential. A series of objectives will be pursued in this project: 1. Refine methods for the isolation of rat mammary MTLn3 and MTC cells from blood. Further develop robust molecular mass signatures for these cell types using principal component analysis. 2. Prepare metastatic tumors using MTLn3 and MTC in the Fisher 344 rat; recover and quantify circulating cells from blood at several time-points. 3. Use the methods developed to analyze blood samples from up to 20 women with metastatic breast cancer obtained through the UC Davis Cancer Center Specimen Repository.

We will further develop a centrifugal separation method for CTCs based on cell density and compare it to a commercial method that isolates cells using immunomagnetic beads with antibodies to epithelial cell proteins. Quantifying the recovery of cells spiked into blood will prepare us for experiments in rats implanted with xenografts to produce mammary tumors, which will shed cells into the circulation. Blood from groups of rats will be harvested at several time points, quantifying the threshold of detection of CTCs and correlating presence of cells with metastases. A major milestone will be detecting CTCs in blood samples from up to 20 women known to have metastatic breast cancer. We will begin to establish “mass spectral” criteria characteristic of metastases, and apply multivariate statistical analysis to the many cells recovered from the blood to identify metastatic cells. These experimental steps will allow us to develop a robust and reliable method for identifying metastatic breast cells circulating in peripheral blood.

To this point our data demonstrates that it is possible to differentiate cells by their subtle chemical differences in disease potential. Our method will have the power of proteomic analysis with the advantage of single cell detection and could be used as the basis for a blood test for breast cancer metastasis.


Final Report (2008)
Diagnosing and treating metastases is critical to the survival of breast cancer patients. Currently, the presence of cancer in the lymph nodes remains the most valuable predictive indicator for metastatic breast cancer; however, as many as 30% of patients considered to be lymph node negative eventually become sick again. This makes it clear that better methods are needed to identify breast cancer metastases. Because breast cancer tumors have been shown to shed cancer cells into the circulation as they become invasive, it has been proposed that detection of circulating tumor cells (CTCs) in the blood provides a way to quantify the risk for developing malignancies. These cells may also help us detect recurrence of the disease.

Our project was designed to develop a novel mass spectrometry (an analytical technique that identifies the chemical composition based on the mass-to-charge ratio of charged particles) method to detect and identify these circulating metastatic cells. There were three aims:

In our project, experiments in Aims 2 and 3 depended upon successfully completing Aim 1, because samples acquired in the latter aims are extremely precious and must be analyzed with reliable, robust isolation and mass spectrometry methods. Unfortunately, creating this truly efficient cell isolation technique did not materialize. Although commercially available methods to isolate CTCs exist, these methods did not produce adequate recoveries of cells from human blood. Our method must be sensitive enough to detect as few as 10 CTCs in 15 milliliters of human blood and it must be efficient and reliable enough not to miss any of those cells.

Although we were not able to move to the animal or human blood experiments, we made significant progress on refining and understanding other steps that will be necessary to effectively characterize CTCs in the future. We have established that the spectra of individual breast cancer cells that are models for different kinds of breast tumors can be differentiated using statistical techniques. We have developed a method to prepare single cells for analysis after the cells are isolated from blood, ensuring that the cellular characterization is consistent and robust. We have established that the mass spectra are the same across time and across samples and have shown that our isolation handling techniques do not affect the chemical signatures of the cells. We have implemented advanced statistical methods that will allow us to identify cells with metastatic potential. Four peer-reviewed manuscripts have resulted from these efforts.

Finally, we have thoroughly evaluated and modified two commercially available methods for cell isolation so that we now consistently achieve 40-50% recovery of “spiked” cells. The ultimate goal of this project, using CTCs as markers for quantifying risk of breast cancer metastasis, remains an important problem. Although our progress was slowed because of the need to spend so much time trying to make the isolation method work, we feel that the ToF-SIMS characterization method will be relevant for analysis of CTCs in the future. We will continue to monitor methods for cell isolation and when these become more robust, we will be well-poised to apply the sample preparation and statistical methods that were developed in this project to quickly solving this problem. Using imaging mass spectrometry to analyze CTCs is a novel application that, to our knowledge, is only being pursued in our laboratory. The valuable support provided by CBCRP for this project is unique to our laboratory and to other funding agencies.

Preparation of single cells for imaging/profiling mass spectrometry.
Periodical:Journal of the American Society of Mass Spectrometry
Index Medicus: J Amer Soc Mass Spectrom
Authors: Berman ES, Fortson SL, Checchi KD, Wu L, Felton JS, Wu KJ, Kulp KS
Yr: 2008 Vol: 19 Nbr: 8 Abs: Pg:1230-6.

Chemical and biological differentiation of three human breast cancer cell types using time-of-flight secondary ion mass spectrometry.
Periodical:Analytical Chemistry
Index Medicus: Anal Chem
Authors: Kulp KS, Berman ES, Knize MG, Shattuck DL, Nelson EJ, Wu L, Montgomery JL, Felton JS, Wu K
Yr: 2006 Vol: 78 Nbr: 11 Abs: Pg:3651-8

Imaging and differentiation of mouse embryos by ToF-SIMS
Periodical:International Journal of Mass Spectrometry
Index Medicus: Int J Mass Spect
Authors: Wu L, Lu X, Kulp KS, et al.
Yr: 2006 Vol: 260 Nbr: Abs: Pg:137-145