Scientists have shown that there is considerable transcriptional diversity between individual circulating tumor cells (CTCs) found in the blood of a single cancer patient, indicating that multiple treatments may be required to adequately kill off a patient’s cancer. Researchers led by a team at Stanford University used an immunomagnetic enrichment device technique known as MagSweeper to isolate CTCs from blood samples taken from breast cancer patients. The investigators then used microfluidic-based single-cell transcriptional profiling to look at the expression of 87 cancer-associated and reference genes in individual CTCs.
The results, reported in PLoS One, showed that there was significant heterogeneity in the transcription profiles of CTCs released into the blood, even among those from a single patient. Importantly, the expression profiles of all CTCs were very different from those of seven different primary and metastatic human breast cancer cell lines that are widely used for drug discovery and in vitro drug testing. “Our finding suggests that perhaps they’re not that helpful as models of spreading cancers,” claims lead investigator Stefanie S. Jeffrey, M.D., professor of surgery and chief of surgical oncology research at Stanford’s Department of Surgery.
The researchers describe their findings in a paper titled “Single cell profiling of circulating tumor cells: transcriptional heterogeneity and diversity from breast cancer cell lines.”
The treatment of metastatic cancer is determined largely by biomarkers from the primary tumor, despite the fact that metastatic tumors and primary tumors may be very different. Given that metastases are seeded at distal sites by CTCs released into the bloodstream, a method for accurately profiling such cells could provide valuable information for guiding therapy, the authors note. However, accurate characterization of CTCs is hampered both due to contamination with white blood cells and the limited sensitivity of analytical methods, which means analyses may have to carried out on pooled CTCs.
To demonstrate that single-cell transcription profiling is possible for CTCs, the team used the MagSweeper enrichment technology to isolate CTCs from breast cancer patient blood samples, and then applied microfluidic-based single-cell transcriptional profiling technique to determine the expression levels of 87 genes in individual CTCs. Blood samples were taken from 20 breast cancer patients with primary tumors, and 30 patients with metastatic disease. Of the 87 genes evaluated, 31 were consistently detected in at least 15% of the CTCs analyzed. Apart from three reference genes that were expressed in all the CTCs, the 28 remaining genes most commonly expressed were found to be relevant to nine basic functional categories: epithelial phenotype, epithelial mesenchymal transition, metastasis, PI3K/AKT/mTOR pathway, apoptosis, cell proliferation, DNA repair, cell metabolism, and stem cell phenotype.
Clustering analyses indicated that while the transcription profiles of CTCs varied considerably, they fell into two broad groups, both of which could be present in the same patient. Cluster I represented a relatively small proportion of the CTCs which, in comparison with cluster II, were characterized by stronger expression of S100A9, CD24, VIM, CXCR4, MAPK14, AKT2, PIK3R1, CTNNB1, CD44, and ZEB2. Importantly, both clusters robustly expressed a number of metastasis-associated genes.
Of particular note was the finding that the CTCs were almost exclusively triple negative (i.e., lacking Er, PR, and Her2 expression), irrespective of whether the patient’s primary tumor was triple negative or not. “Loss of expression of Er/PR/Her2 in CTCs noted in our particular patient sample could explain why therapies that target these biomarkers may fail to control end-stage disease,” professor Jeffrey et al point out.
And surprisingly, all the CTC transcription profiles differed greatly from those of three commonly used primary and four metastatic human breast cancer cell lines. “Overall, expression patterns of <10% (2/28) common tumor-associated gene profiles of CTCs were recapitulated by a subset of tumor cell line models,” the investigators write.
“Although CTC heterogeneity between patients is well recognized, an important finding in our study was that individual CTCs did not cluster by patient or disease stage (primary cancer vs. metastatic cancer), which again supports the concept that these cells belong to subpopulations with phenotypes fundamentally different from pooled tumor tissue,” they conclude. Studying and phenotyping the primary tumor alone may lead to suboptimal treatment selection ... Our finding of CTC variability is consistent with primary and metastatic tumor heterogeneity and suggests that single cell phenotyping of CTCs is a practical approach to exploit this variability for the effective implementation of molecular guided cancer therapy on a more comprehensive scale than possible with mutational analysis of a few known genes.”