SLIDES & TRANSCRIPTS
Saturday, December 6, 2003

Highlights of GU SPORES

William Ellis, M.D.

What we have basically tried to do is make an all-star team from the Northwest to compete with you people back East. We have taken the expertise of the Fred Hutchinson Cancer Research Center , epidemiology and basic science, biostatistics, our expertise at the University of Washington and clinical care and basic science research. The Institute of Systems Biology has an expertise in genomics and proteomics. The group at UBC, University of British Columbia , has an expertise in hormonal therapy in animal models of epitosis, and then we have a collaborative arrangement with Oregon Health Sciences University .

The overall theme of our SPORE is translational research as it relates to prostate cancer progression and metastasis. It does four research projects, five cores, a developmental program project and a career development program.

TOP

The projects are listed here. Paul Lang and Janet Stanford are the principal investigators.

The first project briefly is led by Janet Stanford and David Penson. They are looking at genetic polymorphisms as predictors of tumor progression. They have a cohort of over 800 patients from the early to mid-1990s. They are looking at a series of genetic polymorphisms, as well as some other genetic polymorphisms such as those for Vitamin D and the androgen receptor, to see how these relate to progression.

Project two I am going to talk about more in detail in a little while. Project three is based at the University of British Columbia with Martin Glebe and Colleen Nelson. That is looking at the LNCAP in Shirinogi tumor models and getting expression profiles of those in animals which become antigen resistant or taxane resistant, and then try and develop targets for antisense oligonucleotides.

The fourth project is looking at transcriptome and proteonome stratification of prostate tumor phenotypes, looking at those patterns of expression which predict progression, and also looking at a subset of patients who receive neo-adjuvant, taxine and mitozantrine therapy, and looking at changes in expression profiles.

TOP

I'd briefly like to highlight the second project, which is looking at disseminated prostate cancer cells. What we are trying to do is both describe the phenomenology and also characterize the phenotypes of these cells.

The way we are going about this is isolating the cells and then characterizing them with fairly standard immunohistochemical techniques for antigens of interest. We are working with Lee Hurd and Pete Nelson to perform cDNA micro arrays on the isolated cells. When we find areas of interest, we are going to look to try and confirm these changes in the genome with CGH, and then where we have adequate tissue we are going to correlate the changes both in the tissue and in the circulating cells.

TOP

The micro metastatic problem is that clinically we have a lot of difficulty detecting micro metastatic disease and predicting who is going to fail with definitive local therapy. Our group and others have shown to RT-PCR that tumor cell dissemination is probably a very early event in prostate cancer. These RT-PCR studies all show a trend towards increased tumor cell dissemination in more advanced disease.

TOP

Unfortunately, RT-PCR did not live up to its early promise, and we are unable to predict which patients are going to respond to definitive local therapy ahead of time. It is also limited by the fact that with RT-PCR you get a signal that these cells exist, but you can't actually get hold of the cell to interrogate it.

TOP

So what we need is a method of isolating these cells and also a method of interrogating these cells.

TOP

The way we isolate these cells from the blood and bone marrow, we use a paramagnetic bead system, the Multini Max system, and we use anti-CD-45 and anti-CD-61 antibodies to pull out the lymphocytes and make it keriocytes, and then we use human epithelial antigen to pull out the epithelial cells of interest.

TOP

On this cartoon, when we add the anti-CD-45 and anti-CD-61 particles, we can pull out the lymphocytes and make keriocytes in the magnetic column here. The epithelial cells swell through. We then put an epithelial antibody in to pull out the epithelial cells, allow the rest of the cells to flow through, then we elute the column and isolate the epithelial cells. This gives us a ten to the third to ten to the fourth enrichment of these epithelial cells.

TOP

We can then take FITC labeled antibody to BURF-4, which is another human epithelial antigen, to identify these cells and look at PSA expression by RT-PCR and immunohistochemistry.

TOP

Here is an epithelial cell by light microscopy surrounded by these other lymphocytes. This is fluorescent microscopy.

TOP

Here is a pair of cells right here,

TOP

and this is a cluster of cells by fluorescent microscopy.

TOP

This is PSA staining, showing expression of PSA in a peripheral blood cell.

TOP

This is a cluster of cells isolated from the bone marrow showing PSA expression.

TOP

What is very important to us with this technique is, we are able to use a micro pipette and pull out these cells and pool them, so we can get pools of cells that we can then try and interrogate.

TOP

We don't find cells in all patients, but when we do find cells in the typical patient with clinically localized disease, we can get ten to 20 cells to pool. The patient without evidence of disease is usually less than ten cells; in patients with advanced disease we can usually get 100 or more cells to pool.

TOP

I'm going to briefly review our results in the peripheral blood. Ours are in the bone marrow. We have peripheral blood results which are similar. We find half to two-thirds prevalence of cells in the peripheral blood that we do in the bone marrow.

If you look at our bone marrow results, this is what we see if we do standard RT-PCR for PSA on our unenriched samples. This is the epithelial cells we see in enriched samples, and this is PSA expression in the enriched samples. In our normal controls we tend not to see cells very often, and they usually don't express PSA. In these patients who have elevated PSA but negative histology or are at higher risk for prostate cancer, we start seeing epithelial cells more often, and they do occasionally express PSA.

TOP

In our patients with prostate cancer, patients prior to radical prostatectomy, we see epithelial cells in 75 percent of the patients in the bone marrow. Just over 50 percent of these express PSA. As you look at patients who are treated with hormone therapy, this number goes down a little bit, but in hormone refractory disease most patients will have evidence of epithelial cells and PSA expression.

When they recur, we start seeing these cells again, but the group that is most interesting to us is these patients without evidence of residual disease. In the first year after prostatectomy, 61 percent of the time we see epithelial cells in circulation. As you get out more than a year, that number decreases, it is 47 percent, but we still see it in patients even more than five years out from prostatectomy or without evidence of residual disease.

These numbers are significantly higher than the percentage of patients you would expect to develop a recurrence following radical prostatectomy. Dr. Walsh pointed out yesterday, in the modern PSA era, probably 80 percent of these people are going to end up without evidence of residual disease, because of the low stage that they are in.

TOP

So our conclusions are that we can isolate epithelial cells from the bone marrow and peripheral blood of a significant number of men, both before and most importantly, after radical prostatectomy. We find these cells more frequently in the bone marrow than we do in the peripheral blood, and not all epithelial cells produce PSA. We think this may be due to some loss of the stromal epithelial interaction, something similar to what you see in cell culture, where they make much more PSA when they are incubated with stromal cells or stromal conditioned media.

TOP

We think that these disseminated cells are likely malignant based on the distribution of the cells. There are more cells present in more advanced disease, and also the early cytogenetic analysis.

TOP

We don't know what these disseminated cells are. They may be terminally differentiated cells which can't proliferate. They may be cells that can't survive long term outside the prostate micro environment. They may be cells that are cleared by the immune system, or they may be cells in a long term state of dormancy.

TOP

So our future studies are basically trying to characterize these tumors further through the gene arrays and the other analyses I mentioned at the start.

TOP

I'd like to thank our collaborators in this study as well.

TOP