SLIDES & TRANSCRIPTS
Friday, December 13, 2002

Markers for Bladder Cancer: Where Are We? Where Are We Going?

Ralph deVere White, M.D.

What I want to talk about is the role of markers in TCC. I think you have to look at this in selected clinical situations. I think it's worth remembering that 8 percent of our tumors present at T2 or higher, and they account for 85 percent of those people who die of transitional cell carcinoma. So if we want to make an impact in this disease, we have to look at the T2s, not the 92 percent of Ta, T1s, which account for only 15 percent of transitional cell deaths.

Now, we can do this. We're going to look for markers either by doing tumor analysis or serum analysis. So let's talk about tumor analysis, and I'm leaving urine out of this. We know from the SWOG study and the MD Anderson that if you take people with T2 disease, you give them a TUR, you give them chemotherapy, and you take out their bladder, 40 percent of them will have no tumor, and they will have an 80 percent 5-year survival.

For the patients who had chemotherapy, and had positive nodes, that's 22 percent. Their 5-year survival is only going to be 12 percent. We want markers that will allow us to predict this up front.

TOP

When we look at serum, we've got a different issue. Can we find markers either that will tell you have metastatic disease before surgery? But probably even more important, after you have removed the bladder, or removed the cancer, have we markers that will say they still have residual disease, and you could give adjuvant chemotherapy?

Or if you had a marker that went away, could we then follow a marker that would allow us to say you have early failure, and we give you chemotherapy before you develop the disease? The definition that I'm using for a marker is that it is in routine clinical use.

TOP

The definition that I'm using for a marker is that it is in routine clinical use.

So let's go back to Ta/T1 disease.

TOP

You have the analysis of the tumor stage and grade, CIS. We have response for vesicle therapy. And let me just say we have no other marker in routine clinical use, despite 20 years of study.

TOP

This is the paper from Dr. Heney, written in 1983. And we have not advanced in terms of making predictions of outcome in Ta/T1 disease.

TOP

Now, is it because no other marker works? No. I'm going to show you three papers, all who looked at S-phase. This is a SWOG study of 244 patients given BCG or myelomycin-C, multivariate analysis. S-phase gave additional information to tumor grade with regards to progression. And this we published two years ago.

TOP

Here are 207 patients with Ta/T1 disease. And in this case, S-phase in multivariate analysis was the only predictor of outcome. It outdid grade.

TOP

And if you don't like flow cytometry, and most of us don't any longer -- it uses too much tissue -- here is an immunohistochemistry S-phase analysis using KI-67. Once again, the only predictor of regression is S-phase.

TOP

So what I'm showing you is 455 Ta/T1 patients, three different studies, the S-phase is measured in two different manners, put through multivariate analysis. S-phase is predictive of clinical outcome. Yet, we do not use S-phase. So the problem is we still treat the patients based on the clinical parameters we heard about earlier.

TOP

So the second marker that obviously fits into this is p53. This is Carlos Cardon-Cardo's paper with Harry, published now in 1994, so it is nine years old. If you look at it, p53 is down by IHC. If you were positive, mutative, T1, 76 percent of those went on to get invasion. If you were negative, only 17 percent did, a big separation.

TOP

The next year, this is a paper from England. If you were p53 negative, 1 of 55 patients progressed. If you were positive, 12 of 14. So you would say, gosh, this really looks good. Why aren't we using it?

TOP

Well, here is a review from this year in European Urology. So, IHC was used to judge the status of p53. Is it prognostic? Four papers, 505 patients said, oh, yes it is. Then is it not prognostic? Four hundred and one patients, three different studies, and multivariate analysis.

So the question is why this huge difference?

TOP

TOP

I think there were two reasons. One is problems in methods. And the other is difference in interpretation. Now, can we overcome them?

TOP

This is a study that we published with Cora Sternberg last year in European Urology. These were 40 patients that Vito Pansadoro resected, and Cora had given chemotherapy to, and we did the analysis (Eur Urol 2001 Apr;39(4):430-7).

I just want to show you, 92 percent of those by IHC were p53 positive. In our own patients, it was 50 percent. Clearly, they are not all mutated. So what's the problem?

TOP

We have moved now to doing a functional yeast assay. Richard Iggo was very kind and sent us his assay a couple of years ago from Switzerland. So what happens is you take the yeast. You put in a single allele, and that allele, if it is wild type, drives transactivation, so adenine cyclase is activated and the adenine in the medium is digested. You get this nice big, white colonies.

If it is a partial loss of function, it is partially driven, and you get the nice pink colonies. And if there is mutant or no activation, you get these small red colonies.

TOP

Now, this I want to say is ongoing work. But what I want you to see is this is all fresh bladder tumors. And here are the yeast results. What I want to show you is that all ten have got mutations, two are partial mutations. Three colonies of which were sent, and they all have sequenced mutant. This is mutant p53.

Here is the IHC -- negative, negative, negative. So 3 of 10, the IHC does not tell you the status of p53. One the flip side, 10 are wild type by yeast, 4 are positive by IHC. I think IHC is good, I just do not think it is accurate enough in the individual patient.

TOP

The next marker that really a considerable amount of work has been done is RB, the allotumor suppressor gene. It's been done mainly in at MD Anderson and Memorial. And this Bart's paper. If both were normal, none progressed; 1, 25 percent progressed; both, 80 percent progressed.

TOP

And indeed, this is the very nice paper from Newcastle. It looked at three-year survival in radiation therapy. If you were P21-positive it was 82 percent; P21-negative was 12 percent. Now, why the difference? Well, you can argue this both ways. If you have high P21, you're going to cell or cycle arrest, cells going to apoptosis. The other side of it is that you get time for DNA repair. But it does look good.

TOP

In Cora's patients there was not a single patient that was P21-negative, RB-negative that did survive. But if you had both of those positive, 70 percent did.

TOP

So what's the goal in the T2 TCC? We need marker validation. We need prediction of response to therapy. We need marker discovery. And we need to develop new therapies.

TOP

TOP

Looking at serum markers, this has been looked at now in three different ways, and this is from Paul Gumerlock's work in SWOG. What this is showing is wild type RAS, mutant RAS. And so what you are looking at is it's in the tumor, it's in the plasma.

TOP

Here is a patient with mutant RAS, given chemotherapy; nothing happens. This patient was given chemotherapy, and it goes away.

TOP

And this is Peter Danenberg's work from USC. And he is looking at APC, and this is in lung cancer. What you see is it never comes up, and the patient does not recur.

Here is a patient the day after surgery. The marker comes up, and it takes ten months for clinical failure.

TOP

So TCC should be absolutely ideal. It's high prevalence, easy to get tumor, obtain repetitive tumors, obtain tumors pre- and post-therapy. It is relatively homogeneous, and has a relatively short natural history.

TOP

We should be able to do tissue and microarrays. I think the only way to do this is prospective clinical trials. There is an NCI marker group called PACT, and they have picked three tumors as demonstration models for marker validation, marker discovery. One of those is muscle invasive bladder cancer.

The second thing is the SWOG study, S0219.

TOP

Now, this is a study that I am glad to say, will be activated January 1, 2003. This is a follow-on to 8710 or what Harry showed you. In that study, you got a resection. You got your MVAC. You got a cystectomy. The question is, in those 40 percent of patients with 85 percent 5-years survival need that cystectomy?

So the only difference is in this trial, you would get re-resected after your chemotherapy. And if you are positive, you get a cystectomy. And if you are negative, the patient and you get to choose. It is not a randomization. But what is necessary in this study, to get on the study, you must send the paraffin blocks, and if possible fresh tissue.

So we will have 98 patients prospectively studied before and after chemotherapy. All will have tissue arrays, and a number will have microarrays. And hopefully in this way, we can do both marker validation, marker discovery, and hopefully marker discovery will lead to new treatments.

Thank you very much.

TOP