TOPICS > Science

Extended Interview: Dr. Nicole Urban Explains Clues of Cancer

March 28, 2007 at 5:25 PM EDT

SUSAN DENTZER: What is the Holy Grail likely to be in biomarker research?

DR NICOLE URBAN: Well, the Holy Grail would be to identify cancer using some test. A blood test is — desirable because it’s relatively non-invasive. But, what we would really like to do is to identify the cancer before it becomes invasive so that surgery could be performed preventing ovarian cancer.

An example of that is the pap smear for cervical cancer– which has reduced mortality by 70 percent of more in this country because the pap smear is able to identify a precursor condition– for the cancer. Now, in ovarian cancer, we haven’t identified a precursor lesion which makes it more difficult to identify it. Nevertheless, that would be the Holy Grail would be to identify it before it becomes invasive. The challenge is–

SUSAN DENTZER: So the Holy Grail would be to find disease at its earliest stages before it has actually invaded an organ or spread to other organs.

DR NICOLE URBAN: That’s right. Now, more realistically– we will be happy if we can discover ways to identify the cancer after it has become invasive, but before it– it has become metastatic. So, we would like to identify cancer that can be cured.

To find a panel of markers that would be able to identify ovarian cancer while it is still curable. And, the challenge there is that there may be a very short period of time– from the time that it becomes invasive to the time that it becomes incurable. And, for that reason, what we would like best would be to identify it before it becomes invasive, because then it can be cured by surgery alone, and, does not require chemotherapy.

SUSAN DENTZER: Now, let’s talk a bit about ovarian cancer and it’s a particularly desirable cancer in which to find a very useful biomarker or panel of biomarkers.

DR NICOLE URBAN: Well, ovarian cancer– occurs deep inside a woman in her ovaries. And, there is no apparent sign of ovarian cancer. It does not cause symptoms that are specific to ovarian cancer.

I would not say there are no symptoms. There are some recent– evidence that there may be symptoms. But, they probably don’t occur until the cancer is incurable.

Symptoms of cancer

SUSAN DENTZER: And, some of those symptoms would include?

DR NICOLE URBAN: Well, the-- there are a number of symptoms. Pain is the most obvious. Bloating. Bloating is a very common symptom of ovarian cancer. Frequent urination. But, these tend to occur when there is a mass that is sizable, and therefore, causing these symptoms. And, we want to find the cancer earlier than that.

SUSAN DENTZER: And these are symptoms that women experience commonly.

DR NICOLE URBAN: That's right. That's right. So, when a woman goes to her doctor and reports those symptoms, frequently, the doctor does not think about ovarian cancer at all. He thinks about other kinds of problems, more with the digestive system.

So frequently, it will take six months for a woman to finally be diagnosed with ovarian cancer after she's been worked up for other types of conditions. So, it's a challenge. So, a blood test is particularly likely to be a solution to the problem of ovarian cancer.

So, imaging is another possibility. So, there is trans-vaginal ultrasound. And, it is used. It's being studied-- particularly at the University of Kentucky. And, it's being used in two randomized controlled trials. One in the UK, and one here in the United States.

But again, by the time you see a mass-- the disease may not be curable. So, what we're hoping for is a blood test that would give a signal early enough in the course of the disease that it could be cured, at least with chemotherapy combined with surgery. And, if we could get it early enough, maybe by surgery alone, which I would like to see 'cause chemotherapy is tough.

SUSAN DENTZER: And statistically, when do we tend to find ovarian cancer? At what stage? And what is the result in mortality based on that?

DR NICOLE URBAN: Well, unfortunately-- about 75 percent of ovarian cancer is found in what's called Stage Three, which means that it has spread beyond the ovary, beyond the pelvis, and into the peritoneal cavity and the organs there. And, the cure rates are quite dismal there. Twenty-five, 30 percent at best.

Whereas, if it's detected early-- 90 percent of women survive at least five years, most of them ten years. There are different types of ovarian cancer. So, it does depend on the cell type. But, even the most aggressive cell type, does quite well when it's detected when it's confined to the ovaries. So, ten year survival rates are no worse than 60 percent.

SUSAN DENTZER: And overall, mortality in ovarian cancer has been pretty stable, whereas in other cancers, mortality has declined.

DR NICOLE URBAN: Yes. That's true. And-- the most obvious case of a decline is in cervical cancer. So, cervical cancer, the advent of the pap smear was many years ago, in the 1930s. And, over the course of the last century-- cervical cancer mortality rates have dropped 70 percent.

And, when we just look at what's happening to the mortality rates in breast cancer, they are dropping, I suspect, by as much as ten percent before much longer.

So these are success stories for early detection. And, we would like to do the same thing for ovarian cancer. And we have seen no progress at all. They're just flat.

Known biomarkers

SUSAN DENTZER: Now, we have several known biomarkers now of ovarian cancer. The first of which is CA125. Cancer antigen 125. First of all, tell us about what CA125 is. And then, how good a biomarker it is or is not.

DR NICOLE URBAN: Well, actually CA125 is a terrific marker actually. It detects about 80 percent of cancer. And so, at the time that a diagnosis is made when a woman is symptomatic and has late stage cancer-- about 80 percent of those women do have elevated CA125.

The problem is that it doesn't elevate until the cancer's quite advanced. So, it-- although it works very well, it works extremely well for identifying recurrence. So, it's used to follow women who've already been diagnosed to see if they-- if the cancer's coming back.

But, it does not work well enough for early detection in most people's opinion. Partly because it doesn't detect early stage disease. And, partly because it is positive in a lot of women who do not have cancer. It's not really specific to ovarian cancer.

SUSAN DENTZER: And, why is that?

DR NICOLE URBAN: Well, I don't really know the answer to that question. It could be that it's a marker of inflammation. And, there are other reasons for inflammation.

CA125 is a useful marker though. I do not want to give the impression that it's not useful. And, in fact, there are people who are studying different ways to use CA125 to make it more useful.

Currently, the threshold for positivity in CA125 is considered to be 35 generally, maybe as low as 30 in post-menopausal women. But, there are ways to use the marker, like PSA-- it's similar-- that work better.

So, for example, CA125 varies a lot among women. It varies much more among women than it does within an individual woman over time. So, that 90 percent of the variance in CA125 is across women rather than within a woman.

That makes it possible to establish a baseline for a woman. And then, to ask the question: Is this woman CA125 higher than we would expect it to be for her? And, this allows a lowering of the threshold since most women have a level around ten or 12. If it were to go up to 20, that would be unusual for that woman. And, by calling the signal 20, instead of 35, we can detect the disease earlier because it rises. As the disease progresses, we assume. We don't actually know that. (LAUGHTER) But, it-- there are studies ongoing to test this hypothesis that a longitudinal algorithm-- can detect the disease sooner. And, do a better job of detecting it early.

SUSAN DENTZER: Let's take a step back and tell people what CA125 actually is.

DR NICOLE URBAN: Well, it's a marker. It's a circulating antigen, so it's a protein. It's a protein that has been secreted into the bloodstream. And, you can measure it using a test which involves two antibodies.

Sort of the idea is that the-- the body manufactures antibodies. The immune system manufactures antibodies that bind to antigens. This is the way that-- it works in people and in mice. And so, it's possible to develop a test like this because producing the antigen of interest. And, immunizing mice, females, with this antigen.

And then, the mouse will produce antibodies. And then, the researchers harvests those antibodies. And, tests them to see if they will bind to the human antigen. And, if they will, then if the researcher's fortunate enough that he can find two that binds to different places on the antigen, then he can develop an assay called a reagent.

And, it's very good type of test. It's very specific because if you only use one antibody, it-- that particular antibody might-- bind to some-- promiscuously to some additional antigens, in addition to the one of interest. But, if you have two antibodies-- to a single antigen, the probability that you would pick up the wrong antigen by mistake is very low.

So, the idea is that you find these two antibodies. And, this allows you to measure the-- how much of the antigen is circulating.

What produces CA125?

SUSAN DENTZER: What do we think is producing the CA125? Is it an actual gene changing in a woman with ovarian cancer that is causing the elevation of the protein? Is it an over expression of the-- of the gene itself? What do we think is going on?

DR NICOLE URBAN: You know, I don't think that we know that. CA125 has been around for a long time. Something like 25 years, since '83. And, very little is really known about it.

It's kind of remarkable in a way how little is understood about CA125 considering how long it's been around.

SUSAN DENTZER: So, what is-- what is a-- an appropriate analogy? This is like-- we're back in Ancient Greece. And, we look at the stars. And, we see stars we don't understand that these are bodies out in the-- in space that are like the sun. Is that-- where we are?

DR NICOLE URBAN: Yes. We have just scratched the surface. You know, we have-- very far to go in understanding CA125 in particular, but, biomarkers more generally.

SUSAN DENTZER: Now, you mention another biomarker. What is mesothelium?

DR NICOLE URBAN: Mesothelium is a protein. In the presence of cancer, it appears to be cleaved. And, becomes soluble. And-- then, circulates in the blood stream. So, that when you can measure mesothelium in the blood, that suggests that there's something-- wrong. Mesothelium is secreted when cancer is present in the case of ovarian cancer.

It is also a marker for mesothelioma, which is a form of lung cancer that people who are exposed to asbestos get.

SUSAN DENTZER: You've also looked at a composite marker. Let's talk about that, of CA125.

DR NICOLE URBAN: So, we have looked-- our goal is to put together a panel of markers. So, we start with CA125. And then, we add additional markers to that.

So, we started with some that we had identified ourselves. And, we put mesothelium was the first one. And, HE4 was the second that we identified ourselves. But, we felt that there was-- it was unlikely that the markers that we had come up with would be sufficient. So, we invited collaborators from all around the country who had markers to test their markers on our blood samples.

So, we sent them blinded blood samples. And, they measured their markers. And then, we tried to put together a panel. So, the first one that we looked at was mesothelium because it was the first one that we had.

And, it did appear to compliment CA125. And, by that I mean that some of the cancers that CA125 was missing were i-- were picked up by mesothelium so that if you used a decision rule that said if either CA125 or mesothelium is elevated, then we'll call this test positive. Then-- mesothelium added something to CA125 in a panel. And, that's kind of the criteria that we use in identifying all of these-- markers to see if they contribute to the panel.

The problem with what we've been doing is that we had been doing that with samples that were obtained at the time of diagnosis. And, because of that, most of these-- women have advanced cancer.

So, mesothelium does work quite well as a marker in advanced cancer, same as CA125. What we need is some markers that will identify the cancer earlier. And, I'm hopeful that HE4 will do that. HE4 is our top marker. It's the one for which we have the most hope.

SUSAN DENTZER: So, what is HE4?

DR NICOLE URBAN: It's a gene normally expressed in the male testes. It's a good marker because unlike CA125 or mesothelium that could be expressed in the absence of cancer, HE4, it's unlikely that it would be circulating in a woman's blood unless there were something wrong. So, it has been looked at by a number of people at Harvard, at Dana Farber, at MD Anderson-- here in Seattle.

And, consistently it's ranked the-- the top-- marker for ovarian cancer-- which is encouraging. Now, does not guarantee that it will pick the disease up early. So, I cannot say that we have a-- any evidence that it does.

But, it-- to the extent that we have looked at early stage cases, it seems to perform well in early stage cases that are symptomatic, okay? So, what we really need to do is look at blood samples that were obtained a year or two before the cancer was diagnosed. And, we're only beginning to do those studies.

We are planning to do those studies. So, we have samples like that from a number of (sources, including a colon and ovary trial. This is a trial that's being conducted in the United States that 65,000 women are participating. They're being-- tested for-- in addition to ovarian cancer-- also lung cancer and colon cancer. So, they have taken blood samples from them every year for six years. And then, about 100 of them had developed ovarian cancer.

And, because it's a screening trial, they have intervened. So, the women-- the-- that are being diagnosed on the basis of screening tests. So, the tests that they've been including are CA125 and-- trans-vaginal ultrasound.

But, what it will allow us to do is to have women who are diagnosed at early stages, earlier than would normally be the case because a lot of them were picked up by screening. And, some that the screening missed. Quite a few unfortunately, that that screening missed. But, we'll know-- the state of the cancer at the time that it was diagnosed. And then, we'll have blood samples that were obtained then and a year before, and a year before, and the year before that.

So, we can take-- look at all of our markers by measuring in those blood samples all of our markers. We can see what's called the lead time. So, how early in the course of the disease the marker can pick up the cancer. So, we're especially interested in that five-year period prior to diagnosis. So, we are hopeful that one or more of these markers will give signal two, three, or four years early.

The presence of HE4 in women

SUSAN DENTZER: Now, what is our theory if we-- assuming that we have one-- on why this gene and protein normally expressed in the testes of males would-- would end up in a female and play a role in ovarian cancer?

DR NICOLE URBAN: What a good question. I don't have a theory. I can tell you that HE4 is expressed by the fallopian tube. And, there is a possibility that the fallopian-- that cancers originate in the fallopian tube. The serious cancers. The most aggressive types. It's not impossible that they originate in the fallopian tubes. And, that the HE4 is expressed there.

SUSAN DENTZER: And then, travel up to the ovary?

DR NICOLE URBAN: Mm-hmm. Actually, the theory is that they-- it might-- that there may be precursor lesions in the fallopian tubes. And, that they may essentially spread through the ovary to inclusion cysts in the ovary. That's a theory. It's not proven.

No one knows. I mean, it's-- it's unfortunate that so little is understood about ovarian cancer and how it begins. So, there isn't-- there isn't really even a theory about it. There are a number of different theories about it.

SUSAN DENTZER: So, it might turn out some day we'll call this fallopian tube cancer.

DR NICOLE URBAN: Yeah, it's very interesting. The prophylactic oophorectomies that are being performed in the women who are at very high risk, who are-- mutation carriers-- the occult cancers that they find in those women, about a third of them are in the fallopian tubes. And, about a third of them are in the ovaries. And, about a third of them are primary cancer, which means that the-- the cancer is there in the perineal cavity without appearing to have begun in the ovaries.

So, it's-- it's a puzzlement. But, how HE4 plays into this? I can't tell you.

SUSAN DENTZER: This just underscores in-- in a broader sense how our sort of organ based taxonomy of cancer is probably going to fall apart in the end.

DR NICOLE URBAN: Yes. I think that's true.

SUSAN DENTZER: We'll call all these things very different things 20 or 30 years from now--

DR NICOLE URBAN: That's right--

SUSAN DENTZER: --and not call them breast cancer, or ovarian cancer, or what have you.

DR NICOLE URBAN: Yes, I think that may well be the case. We'll have a molecular classification.

SUSAN DENTZER: So, back to HE4. What-- why is it such a good biomarker now? Why do we think it's such a-- a good candidate for being almost a-- a more definitive type of biomarker than the ones we're working with presently?

DR NICOLE URBAN: Because it is-- almost-- it-- it's the most specific of the markers that we have. So-- the healthy normal women almost never have circulating HE4. And, even the women who have benign ovarian tumors seldom have HE4 in their bloodstream, which is not the case for some of these other markers.

So, CA125, those tend to be elevated a little with benign ovarian conditions. So-- so, HE4 is the most specific, which the definition of 'specific' is that it's not elevated in other conditions. Only in ovarian cancer. So, when other types of cancer are looked at, or other disease of the ovary, or other gynecologic conditions, HE4 tends not to be elevated, whereas some of the other markers do.

Quite a few of the markers that we have are not specific to ovarian. Whereas, I don't know of any other type of cancer where HE4 is elevated. So, that's its claim to fame. Whether or not it will detect the disease early remains to be seen.