SLIDES & TRANSCRIPTS
Friday, December 13, 2002

Molecular Genetics of Bladder Cancer Clinical Implications

Carlos Cordon-Cardo, M.D., Ph.D.

Based on the natural history and the morphology of these diseases, years ago we already decided to test the hypothesis that these two morphological entities also represent distinct molecular pathways, and that lesions that were of this pathologic nature probably had a very different molecular make-up, and they were the ones that could progress after specific molecular alterations would occur. And that rather the majority of lesions that would progress are the ones that are coming from these early carcinoma in situ.

To make a long story short, the participation of different groups from Peter Jones, to David Sidransky, to Margaret Knowles in our own group using allele typing in earlier studies, and later on the studies conducted by Fred Waldman and Wider Sauter utilizing comparative genomic hybridization and some other techniques has shown up. But these distinct entities, and these distinct pathways are really reflected by different molecular alterations.

Superficial lesions are very much represented by molecular alterations of oncogenes including RAS, as well as some specific growth factors. These are one of them. And it seems like the lesions on a yet to be determined gene on the long arm of chromosome 9 are very much responsible for either very early events, or for the establishment of this disease.

Rather, chromosome 9 deletions and RAS mutations are rare, if present at all, in carcinoma in situ. And now that has been established by different groups. Rather, these lesions are characterized by very early alterations of the two prototypes tumor suppressor genes, p53, which controls health fate and apoptotic mode, and probably that's the most important critical function of p53. And RB alteration, RB being the master switch of cell proliferation.

So if you disrupt cell death and proliferation, you have an aggressive disease. If on top of it you also disrupt cell fate, you are going to have accumulation of other alterations, which is what happens with these lesions.

TOP

Is that true? There are two ways for me to be able to convince you that these pathways are real. One of them is to produce genetic evidence. The other, clinical evidence. Genetic evidence has been produced based on the fact that we have various specific bladder cancer models thanks to the work of Henry Shaw and Zhu Wu at New York University. They identified a gene that is unique for the urothelium, and they called that gene uroplakin.
The uroplakin is a specific gene for the urothelium, and it helps in structuring the urothelial form of translational epithelium. Now, utilizing the uroplakin promoter in order to drive the expression of the RAS of the mutation that we just were discussing, what you produce, it is unique tumors that grow exophytically. And they can be so large, that you can fill up the entire bladder.

But as much as we have looked with Zhu Wu, we have not seen any of these tumors disrupting the lamina propria, and producing invasive disease. So alterations of growth factors of genes such as RAS, they disrupt cell proliferation, and they definitely disrupt some structural events, but they are not sufficient to produce an aggressive, invasive disease.

However, if instead of driving the RAS oncogene, you drive the expression of the large T antigen, an antigen that captures and destroys both p53 and RB, and you drive these expressions specifically to the uroepithelial cells, what you reproduce, it's a carcinoma in situ that is multifocal, that becomes invasive, and in most of the cases, metastatic.

So we believe that this model proves that there are different pathways, and that some of these genes are the genes that are very much involved in the pathogenesis of such diseases.

TOP

How about clinically? Well, we already heard some of the comments from Dr. de Vere White that two tumors that look alike under the microscope, such as these two T1 lesions, may behave very differently. And that was the case actually in these two patients in a study that we published with Dr. Sarki some years ago.

And a single marker in this cohort of early blood or cancer, T1 blood or cancer was able to identify those patients that were going to do well, versus those patients that were going to do poorly. A single marker, at least in the context of this group of patients, was able to establish an important significant difference.

Based on that, protocols have been put into effect, and this is one of the protocols that Weaver Banning has been leading at Memorial in which patients may be selected on the disease that they have, based on p53 status with an endpoint of bladder preservation.

And these are exercises that I think that we need to rethink. We need to go back to the drawing board, and we need to see if they are really helpful, and if we can navigate with biology, having as the background, anatomy. So we want to navigate in the coast of anatomy by introducing and by implementing biology.

TOP

It is clear that these molecules don't work alone. And as we have heard, we need to make the concept of analyzing pathways a clinical reality. Not just a biochemical and a basic science exercise. We know that p53, because of its critical functions, it's very much under the control of a very important molecule, the mdm2, a molecule that will present p53 for degradation with it is not needed, or when it is faulty.
The problem is that mdm2 is in very critical region of the human genome. It is chromosome 12q, a region that is commonly amplified. So mdm2 can take care of p53. So in some cases in which p53 is mutated, and mdm2 may be altered, another important marker that controls the power of mdm2, the so-called p14 arF, which actually resides on the chromosome arm 9p in the 9p21 region, where we have seen a lot deletions.

So when we analyzed very recently, the power, not only of one marker, but of the pathway, it's quite clear that when all of them are in wild type confirmation, you get a much better biologically defined disease. But when one or several of them are altered, and when you have the entire pathway altered, as you can see, you are doing much more poorly.

So maybe we should not concentrate in looking at single markers. Maybe you should be able to identify downstream and upstream molecules in these pathways that can help and be more precise about the nature of the molecular changes that we are studying. And molecular pathology doesn't mean molecular genetics. Sometimes techniques such as immunohistochemistry or in situ hybridization may be as powerful as conducting a microchip analysis.

TOP

So if all of these things are so important, why are we not using them? And this is the question that most of us are asking ourselves. We are not using the power of biology to navigate into pathology and into the anatomy, because still there are lots of problems that we need to overcome as a group of professional individuals.
There are important differences in technical approaches. There are differences also in classification schemes, and quite a bit of interobserver variability not only in how we are going to set up the criteria for biology, but also in how histopathology is being read.

And finally, most of the studies are conducted with rather small cohorts of patients. So it's time to attack these limitations, because if we, as a community of clinicians and clinical scientists want to produce something that is going to move beyond the laboratory, we need to address that. And probably the only way to address that is if we have a system that can help us in moving from one step to the other.

TOP

People in industry have learned that in order to bring the drug from the clinical discovery to a standard of care, that drug needs to undergo specific phases. And they are very well defined: Phase I deals with toxicity, Phase II with dose escalation, Phase III, if this new drug will improve treatment compared to current therapies, and Phase IV, is the final validation utilizing a multi-institutional validation trial approach.

Probably we should do the same with markers. If we don't have a way to define where the marker is, we are all going to fight constantly, because we are not going to have any criteria for which to assess the clinical relevance of what we are doing. So if we define a new marker that is of interest, maybe phase I should be a pilot study.
An assay should be developed, and everybody agrees and everybody follows that same assay in order to then produce the prospective clinical analysis and prospective clinical trials that will allow us to have enough data in which to base our scientific hypotheses in the clinical reality.

And finally, I think that we should put apart our intellectual and scientific differences and clinical differences, and we should come around the table to be able to produce this kind of approach. If we would do that -- I think that p53 has seen a good deal of retrospective. We need more prospective clinical trials. And we need to produce multi-institutional validation trials.

TOP

We also need to take into account not only p53, but rather the entire pathway, because as Dr. deVere White pointed out, there are several papers that show how critical is p53 in the context of p21.

Are we far behind? There are not many studies, and with the small cohorts of patients. We need to do more work. And what is more important to us is that based on the fact that new technologies are being developed, there are a bunch of interesting new markers from growth factors to solid lesion molecules to angiogenesis, pathways that need to be further analyzed, because we are going to have very powerful tools for discovery.

But if we don't filter this information of discovery in order to bring some of that into standards of care, this is going to be just an intellectual exercise, and we are all losing our time in here. I would rather be at my laboratory, or at home with my family. So if we don't have a way bring discovery into clinical standards of care, there is a problem in the system.

TOP

Maybe one of the ways to do it is to use these phases, and to use well annotated microarray tissue, tissue that may come from different institutions, read by a group of pathologists that they agree on all of the characteristics of these lesions. And once this is done, to move markers from retrospective studies to prospective studies, and to multi-institution validation studies.

TOP

Let me just show you one of the latest exercises that we have done. As it was said early this morning, we just published a paper on the validation of the utilization of high throughput technology to classify bladder cancer cell lines.

We have done now the same with bladder tumors. We used initially cDNA, but now we are utilizing the affymetrix microchip that incorporates most of the genes that you want to know. Normally, your tumor clusters are very different from superficial and some invasives that cluster together in here, and from invasive and some superficials that are in here.

These patients do quite well. These patients don't do so well. This is not just superficial. It also included some invasive. And this is not just invasive. It also includes some superficial. So the genes that identify these different categories are many.

TOP

One of them is a gene that I didn't know anything about. It's called moesin. And this is an exercise that we will need to do. We will need to learn more. We will learn a lot. There is a lot to be learned. These genes are involved in the maintenance of cell adhesion.

TOP

Superficial lesions are very much positive in the cell surface, but most of the invasive lesions are gone. The statistical significance, it is quite significant when you compare the expression of this gene in superficial and invasive lesions.

TOP

But more important, when you take all of the group of the patients, and you try to stratify with just one new marker, moesin, the presence or the absence of these specific new genes products, it's statistically significant, as much as almost p53, to define the two groups.

So if we can integrate the power of biology with what we have learned in anatomy and the clinical pathology of the diseases that we are trying to analyze, we may be able in the long run, to be have a better assessment of the disease as a process, and of the different stages of the process as well.

TOP

And I just would like to thank Dr. Sanchez-Carbayo specifically, because lots of the work in the microarray is being done now with her and NYU and Dr. Wu for the development of this wonderful new model system utilizing the uroplakin promoter.

TOP