Implications of epithelial plasticity (CK-) and small CTCs on therapeutic response
“Civilization is a progress from an indefinite, incoherent homogeneity toward a definite, coherent heterogeneity.” -- Herbert Spencer, a British philosopher
Similarly, our view of CTCs as misunderstood homogenous cells has evolved as well. We now see that CTCs manifest as heterogeneous populations. At Epic, we are now starting to characterize heterogeneous populations and associate them to clinical outcomes.
We know that CTCs are cells that escape from solid tumors and enter the bloodstream. As we developed the Epic Sciences platform we identified different types of CTCs, including “traditional” CTCs and a host of novel CTC candidate subpopulations (read more about different CTC populations). For example, novel CTC candidates cells include dying cells (apoptotic CTCs), cells which lack epithelial markers and may be undergoing epithelial to mesenchymal transition (EMT), which are notated as “CK-“,CTCs and cells that are the same size or smaller than normal white blood cells as “small” CTCs.
Of particular interest are the CK- and small CTC candidate cells. CK- candidate CTCs, which lack epithelial markers, may be de-differentiated cancer cells (cancer stem cells) which development biologists have described as possessing epithelial plasticity. The potential presence of a cancer stem cell (CSC) could provide value in prognosticating outcome and identifying patients at high risk for metastasis. EMT CTC populations could also explain how tumors evade targeted therapy.
Small candidate CTCs are interesting as it is known that some cancers can evolve from harboring a larger cell pathology at diagnosis but convert to a smaller cell pathology at therapeutic progression (i.e. non-small cell lung cancer can convert to small cell lung cancer). Should a patient have a changing pathology it generally would require a different therapeutic course to provide efficacy. Therefore, Epic’s ability to detect the presence of small CTCs may provide a way to identify if such a conversion is in process with a simple and non-invasive test.
To understand if these “non-traditional” CTC candidates are indeed tumor cells and to better understand the frequency of these cells in contemporary cohorts, we worked with pathologists and oncologists at The Institute of Cancer Research in London and University of Michigan.
We confirmed that these novel CTC subpopulations have cancer related proteins such as androgen receptor (AR) and genomic characteristics (alterations to ERG, PTEN) which are indicative of cancer. The frequency of CTC subtypes in relationship to traditional CTCs was analyzed, and we were able to confirm that many subtypes can be very frequent and have association with aggressive disease.
GU14 Abstract #209: CK- and small nuclear size circulating tumor cell (CTCs) phenotypes in metastatic castration-resistant prostate cancer (mCRPC).
Figure from Abstract #209 identifying a small CTC which has an intact nucleus, CK+, CD45-, AR+, but smaller than surrounding white blood cells (WBCs). These small CTCs also harbor ERG gene rearrangements which in other research has shown to be >99% specific to prostate cancer.
Figure from Abstract #209 graphing CTC nuclear area (y axis) across mCRPC patients. WBCs size distribution of >2M cells seen on right, with single standard deviation (SD) lines drawn. Bottom chart identifies patients’ CTCs based upon size (green are CTCs >1SD larger than WBCs, red are within the SD of WBCs, black are >1SD smaller than WBCs). CTC isolation techniques that utilize filtration or size exclusion would likely miss all CTCs labeled as red and black. Small CTCs have been implicated in patients who are insensitive to androgen receptor targeted therapies.
In collaboration with oncologists at Memorial Sloan-Kettering Cancer Center, we found that some of the CK- CTCs express mesenchymal proteins, indicating that these cells are likely to have high metastatic potential. We compared the frequency of detection of traditional and CK- CTCs by the Epic platform compared to the CellSearch® platform (which enriches for EpCAM+ CTCs) in patients.
Importantly, patients who exhibited a de novo resistance (no or very little therapeutic benefit) to AR therapy such as abiraterone (ZYTIGA®) or enzalutamide (XTANDI®) harbored higher frequencies of CK- CTCs. Platforms which select for epithelial markers like EpCAM, such as the CellSearch® platform, will be unable to detect CK- CTCs in patient samples.
AACR-PCF Abstract #A11: Frequency and characterization of circulating tumor cell (CTC) populations in metastatic castration-resistant prostate cancer (mCRPC)
Figure from Abstract #A11 identifying a CK- CTC. The CTC has an intact nucleus, is CK-, CD45-, but has expression of N-Cadherin (a mesenchymal protein). The CTC also harbors an ERG gene rearrangement (>99% specific to prostate cancer).
Figure from Abstract #A11 highlighting the frequency of Epic Traditional CTCs (dark orange) & CK- CTCs (light orange) compared with a matched blood sample analyzed with CellSearch® (blue line, CellSearch limited to 200 CTC count by processors)
GU14 Abstract #132: Molecular characterization of circulating tumor cells (CTC) and CTC subpopulations in progressive metastatic-castration resistant prostate cancer (mCRPC).
|
Blood Draw |
Epic CK- CTC/7.5mL |
|
True Responders; n=4 |
26 (0-53) |
|
Acquired Resistors; n=19 |
28 (0-150) |
|
de novo Resistors; n=33 |
95 (0-1440) |
Table from Abstract #132 identifying incidence of CK- CTCs in true responders, acquired resistors (patients which have a therapeutic response and then a gradual resistance), and de novo resistance (no or very little therapeutic benefit). As a cohort, patients with a de novo resistance had an increase mean CK- CTC frequency than patients who benefited from androgen receptor targeted therapies.
Summary:
Previously, CTCs were thought to be large cells with an intact nucleus, high epithelial marker expression, in the absence of hematopoietic (blood cell) markers. With the Epic platform we have found that CTCs are heterogeneous, and there are definitive tumor cells which can be the same size as white blood cells or smaller. We have also found that CTCs can have epithelial plasticity and exhibit a mesenchymal protein signature.
Important to our understanding is that these CK- and small CTC populations are common and seen with different frequencies from patient to patient. The presence of small and CK- CTCs was associated with poor diagnostic and prognostic features.
Moving forward, we will continue to examine the morphology and perform protein and genomic characterization of these CTC subtypes as well as their relationship to cancer stem cells and EMT markers. Importantly, it will be of interest to determine their prognostic and predictive value in determining therapeutic benefit for both training and validation in patient cohorts.
Read the next blog post from GU14 + AACR-PCF:
Highly sensitive, single cell analysis: the importance of assessing the heterogeneity of cancer