[gdpawel]

'lizbeth

I think you may be thinking poles and oranges. There are "prognostic" biomarkers and there are "predictive" biomarkers. CellSearch seems to be a good prognostic to monitor treatment response (circulating tumor cells). And this new circulating cell-free DNA carrying tumor-specific alterations (circulating tumor DNA) seems to be more effective at monitoring metastatic breast cancer than current biomarkers of circulating tumor cells, according to a proof-of-concept study. It is a blood test that tracks fragments of DNA shed by dying tumor cells and could one day be used to monitor how well patients are responding to cancer treatment also.

Researchers have been focusing on the development of sensitive assays that allow the specific detection of single tumor cells or small amounts of cell-free tumor DNA in the peripheral blood of cancer patients (Annual Review of Medicine Vol. 63: 199-215). Quantification of circulating DNA by real-time PCR may be a good and simple tool for detection of cancer with a potential to clinical applicability together with other current methods used for monitoring the disease (DNA Cell Biol. 2008 Aug;27(8):415-21).

Doctors hope to use this someday to assess how a patient's tumor responds to treatment. Circulating tumor DNA might give doctors an earlier warning that a drug isn't working. The test might allow patients to stop taking a harsh drug that's not working, sparing them months of side effects. Doctors could then prescribe an alternative drug, although there's no evidence yet that switching drugs sooner would help people live longer.

Again, this test is "prognostic" and not "predictive."

In drug selection, there is a problem with growing or manipulating tumor cells in any way, like they do with CTC testing. When looking for cell-death-related events, which mirror the effect of drugs on living tumors, cells are generally not grown or amplified in any way. The object is occurrence of programmed cell death in cells that come into contact with therapeutic agents.

How do you aggregate a sufficient number of cancer cells to make accurate determinations? Detectable tumor cells in the peripheral blood are present only in extremely small numbers. This precludes allowing a sufficient number of cells to incubate for a few days in the presence of chemotherapeutic agents. Analysis of a relatively small number of isolated cancer cells cannot yield the same quality information as subjecting living cells to chemotherapeutic agents, begging the question of whether or not it can accurately predict which drugs will work and which will not.

CTCs are free-floating cancer cells that can remain in isolation from a tumor for over twenty years. What is the relationship of such long-lasting cells to the tumor cells that need to be attacked through tested substances?

Then there is the question of heterogeneity. The original Immunicon research team really became known for their ability to track and isolate circulating tumor, endothelial, immune and other disease associated circulating cell populations and then using every tool available to further characterize them. The problem they know is the heterogeneity of all these cell populations is greater than any one thought thus defining and characterizing them is more difficult as is finding them - also finding vital ones - as many if not most are dead or dying - this is one of the reasons why the metastatic process is so inefficient.

Tumors in the body are genetically variable. What is the relationship between CTCs and primary tumors or their already established metastases? It has already been established that the gene expression profile of a metastatic lesion can be different compared to that of the primary. The status of the marker Her2/neu in CTCs sometimes differs from that of the original primary tumor, which would point to different prescriptions for Herceptin.

The number of cells discovered in the CTC technique has turned out to be a good prognosticator of how well empiric treatments are working, but less certain in the ability to use it for drug selection. The "problem" is in isolating and analyzing single cancer cells. The supposition is that common cancers can be detected and cured through analysis at a genetic level of a small number of cells or even a single wayward cell.