What we have here in Toronto is a large research registry of BRCA1 and BRCA2 carriers, which began in the 1990s and has now grown to about 7,000 women with these mutations. We have information about their families as well, and we can use this resource to ask questions about the causes of cancer, about screening, prevention, and about different risks associated with different mutations. The registry has grown to become a really invaluable international resource. We have about 50 research hospitals from 11 countries that contribute to it.

“My biggest surprise is that 10 years ago I would have said that our ultimate goal would be to find a great chemoprevention drug and to prevent cancer with a pill. I'm disappointed that that has never materialized.”

This was back when the field of cancer genetics was really just beginning, and I was teamed up with Gilbert Lenoir at IARC in Lyon, France. We set up a lab looking for cancer genes, in collaboration with Henry Lynch, who is really the senior statesman in the field. He spent much of his life collecting data from a large number of families in a registry in Omaha. So the three of us started working on what could be the molecular basis for these cancer families.

We do studies, for instance, on the means of preventing cancer in BRCA mutation carriers. We do studies on screening. One of our notable papers was on the superiority of MRI over mammography in detecting early cancers in this high-risk population. We’ve studied the prevention of cancer in these high-risk women—tamoxifen use and preventive surgeries in particular. Now we’re doing studies on hormone replacement therapy, on treatments and the outcome of surgeries.

My biggest surprise is that 10 years ago I would have said that our ultimate goal would be to find a great chemoprevention drug and to prevent cancer with a pill. I’m disappointed that that has never materialized.

An interesting point is that the BRCA1 gene was mapped in 1990 and we identified it in 1994. BRCA2 was mapped in 1994 and identified in 1995. This was all done using older techniques; we did much of the work with Southern blots on DNA probes. For every marker that we wanted to look at, we had to get a different reagent. It was slow and painstaking work.

But in the last 10 years there has been no new discovery in breast cancer genetics that even comes close to the impact of BRCA1 or BRCA2. No gene has been discovered since 1995 that is equally important to individuals or a population or that has led to the development of a genetic test. That is interesting given that since 1995, we’ve seen the sequencing of the human genome and the completion of highly informative maps for linkage, single nucleotide polymorphisms, and other polymorphisms. We have progressed by orders of magnitude in the last decade, in terms of developing the technology for doing genetic linkage studies, and yet no new genes have been identified. In retrospect, it shows that BRCA1 and BRCA2 really have had unrivaled impact in the world of breast cancer genetics and in the world of cancer genetics at large. It’s highly unlikely that another gene will ever be discovered for cancer susceptibility that has anywhere near the scope of BRCA1 or BRCA2. In the last decade no other genetic test for breast cancer susceptibility entered into the clinical arena.

I don’t think of it that way. It’s very encouraging that we can actually say out loud that we have found the two most important genes for breast and ovarian cancer predisposition. It gives us the boundary and defines the scope of the problem. Personally I’m much more excited now by doing clinical studies on BRCA1 and BRCA2 and actually using these genes to find ways of preventing cancer than I am in doing speculative studies in looking for new cancer genes. To a large extent that’s what we’ve been doing for the past decade.

Many people are now predicting that, using new technologies and sophisticated statistical and laboratory techniques, we’ll be able to really pinpoint multigenetic models that will allow us to individualize predictions of breast cancer risk and the success of treatment. This is supposed to be the next 10 years of the field. I reiterate, though, that the main achievement is identifying BRCA1 and BRCA2, and I’m much more enthusiastic about finding concrete ways of reducing risk in this high-risk population. For example, being able to say that MRI screening works, that we can prevent cancer with preventive surgery and tamoxifen therapy. That’s the kind of research I’m interested in: taking advantage of the information we have in our pocket already about these genes, rather than trying to plan for some utopian future where we’re able to predict, using novel, complicated models, risks of cancer and the individual likelihood of success with treatments.

Well, the risk for breast cancer in a carrier of the BRCA1 is about 80%. We know that can be reduced to a couple of percent by preventive surgery, which is mastectomy. Now our goal is to minimize the cosmetic impact of surgery and yet maximize protection. It’s interesting because the initial idea that one might deal with this high risk of breast cancer by removing breasts was, in many respects, met with horror. This was seen as a mutilating and invasive procedure. In the last 10 years, we’ve seen an enormous change in attitudes toward using preventive surgery, although we’ve seen no change in the women’s reluctance about using tamoxifen, which I consider to be a much less dramatic option than preventive surgery. Still, fewer than 10% of BRCA carriers are taking a drug like tamoxifen now to prevent cancer. In my practice about 40% of the women are having breasts removed when they find they have the mutation. This is why we changed our priority to trying to identify a surgical option that is the best way to preserve body image and the women’s sense of self and at the same time prevent cancer.

It’s the perception of side effects—that’s certainly the reason quoted to us. We’ve asked them and published on this, and they perceive the risk of side effects as a compelling reason to not take it.

This is a very interesting phenomenon, because we don’t necessarily see it the way these women see it. For example, in our database of carriers, we have seen 70 new cases of breast cancer among those women who were healthy when they entered the study, and we’ve seen one new case of endometrial cancer. Tamoxifen would cut the risk of breast cancer in half, but it might double the risk of endometrial cancer. These women don’t consider that a good trade-off, even though the way we see it, we would reduce the number of breast cancers from 70 to 35 and increase the number of endometrial cancers from one to two. They see it as a trade-off of one cancer to another. These kinds of psychological issues are fascinating to me and to some extent are the most intractable. No matter what genes we find, no matter how well we refine our estimates of risk, no matter how well we predict cancer, ultimately a decision has to be made about prevention—either surgery, taking a pill, or going for a screening test. It’s a matter of intense interest to me how these decisions are made. In my mind, they’re often made wrongly, but the women are the ones ultimately responsible for their own health. And often these issues tend to be ignored.

Well, in the world of hereditary cancer, we tend to be dealing with relatively small populations. It’s difficult to propose preventive studies, for example the kind of studies that led to the licensing of tamoxifen for treatment, and now for prevention. The studies that we do tend to be limited to hundreds or maybe a few thousand women, and then randomizing them is really problematic. So far, to my knowledge, there has been no randomized clinical trial done in hereditary cancer. If we can’t do that, we have to make decisions about prevention and treatment with indirect evidence—observational or historical evidence. Many physicians trained in the world of evidence-based medicine are very reluctant to guide their practice on the basis of the kinds of studies we can do, these observational studies. We agree that the randomized controlled trial (RCT) still remains the sin qua non of evidence in medicine, it’s just that we may never see that level of evidence accumulate with such a small hereditary subgroup. We have to use other evidence, but this is a problem. There’s quite a bit of debate in the hereditary cancer world whether or not it’s appropriate even to offer preventive strategies based on the results of nonrandomized studies, even though it’s recognized that data from RCTs may never be forthcoming.

The challenge to cancer prevention is enormous. It’s particularly difficult in breast cancer, because we have no proximal or intermediate marker that really can be used instead of cancer itself. In colon cancer, for instance, people are pretty well served by looking at polyps. If something reduces the risk of polyps, it’s likely to reduce the risk of cancer, although that’s not proven. The people who study cervical cancer are able to look at early neoplastic lesions, and it’s probable that preventing the lesions will prevent cancer. With breast cancer, we have markers, but none are a replacement for cancer itself. So even though we have things like mammographic density, rates of chromosome breakage, or hormone levels, all of which are of interest and seem to be loosely linked to risk, they’re nowhere nearly correlated enough to allow us to make the assumption that if we reduce them, we reduce cancer risk. In the real breast cancer arena, we’ll never be satisfied that a drug is able to prevent cancer unless we actually count the number of cancers in an RCT. That’s an enormous challenge. It seems like, on average, only one such study can be done every 10 years.

Well, I am very fortunate that I have a research chair endowed at the University of Toronto that gives me the opportunity to work full time on breast cancer research. That’s been an enormous support to me. And I’m also fortunate to have many collaborators around the world who are willing to share a lot of data, and allow us to put together the kind of numbers we need to do these studies. And the funding agencies of the Canadian Breast Cancer Foundation and the Canadian Breast Cancer Research alliance have been supportive of us over the past decade.

Steven Narod, MD, FRCPC
Centre for Research in Women’s Health
Toronto, Canada