WEDNESDAY, MARCH 8, 2000

"Studies in Stage III Breast Cancer: What Can We Learn About Biology and Prognosis"

Stage III breast cancer represents a unique clinical situation. It is actually considered to be early stage, that is, before cancer has spread. Stage III cancers have tumors of at least 5 cm. and the presence of some positive nodes. Inflammatory breast cancer is also classed as Stage III. Patients at this stage will usually get some form of therapy prior to surgery. This is usually chemotherapy, but can be hormone. A lump or mass can be measured and followed over time to determine if it is shrinking. The advantage to this versus surgery followed by adjuvant therapy is that you can see if the therapy is working before you have surgery. People with a very good response to therapy tend to do very well, with a small risk of recurring. Treating cancer with neoadjuvant therapy is also a good way to look at new treatments. New therapies, which may be effective, but not curative, in the metastatic setting can improve outcomes and actually save lives. We have been looking at MRI as a way to follow what's happening in the breast before surgery and after neoadjuvant treatment. We can clearly get information about shrinkage of tumor and other biological changes in the breast. A new study is looking at breast tissue after surgery and exposure to therapy by looking at the induction of different genes. Genes encode all the protein in our body. Certain changes in genes (mutations) may cause the cell to change in a way that underlies the development of cancer. Looking at the cancer cell before and after therapy and noting the pattern of affected genes will tell you whether or not a specific therapy is working in a matter of days after treatment.

Can this sort of treatment be done at Stage II? Moving toward tailoring therapy and doing so before surgery can be done in Stage II cancers. Preoperative therapy has traditionally been done with tumors greater than 5 cm. But the trend is to use this therapy with smaller and smaller tumors. Although three large studies have shown that chemotherapy before or after surgery results in the same outcome, neoadjuvant might make a difference in whether or not a patient is a candidate for breast conserving surgery. Most important is the information we receive which tells us we are or are not on the right track with current therapy. The study that needs to be done would randomize two groups: (1) receiving surgery and then chemo, (2)pre-op chemo, if tumor shrinks, continue therapy and then do surgery, or pre-op chemo, if tumor increases, switch therapy and then do surgery. You would then compare the two treatment groups. Currently about 80% of patients have a good response to their therapy. The question is whether you could cull out the remaining 20% with this method.

Is there any neoadjuvant therapy for inflammatory breast cancer, like Herceptin? Currently, we are not using Herceptin in the neoadjuvant setting except in someone who is clearly inoperable, because of potential cardiac toxicity. There will be an NSABP study in early stage breast cancer using Herceptin. People with positive nodes and HER2+ cancer will receive AC followed by Taxol with and without Herceptin. We will have to see whether the benefit of Herceptin outweighs the toxicity. The point is that we must do controlled studies, we cannot just get information from observation to determine whether this is a reasonable strategy. Outcome studies may be more a reflection of their situations than of their treatment. This is illustrated in an analysis of data from women with breast cancer in Finland. They tried to look at the effect of women that had had tamoxifen versus those that didn't. We expected to find what had been shown in clinical trials, that is that using tamoxifen after early stage breast cancer lowered your risk of recurrence by about half. After analyzing the data we found the opposite: the women with tamoxifen had about twice the risk of recurrence and death. This was an uncontrolled (observational) study. The women who were given the tamoxifen had higher risk cancers. Consequently the results had more to do with the population selected to receive the therapy. In these types of studies we cannot normalize for characteristics we don't know, and so might miss the actual cause and effect relationship. Another example of this process is a small Canadian study of transplant candidates for early stage breast cancer. It was found that the potential transplant patients got thoroughly scanned to eliminate those with metastasis, therefore those pre-selected for transplant were much better candidates for the procedure. Transplant patients seemed to be doing much better, but this was erroneous and due exclusively to the selection process. Although randomization is not popular, it does eliminate bias.

Can different tumors respond differently to therapy? The response is the same but the geography is different. Infiltrating ductal carcinomas shrink from a "grape" to a "raisin", while lobular carcinomas shrink from a "grape" to a scattering of "seeds", making this group less a candidate for breast conserving surgery. This is seen in MRIs.

Can vitamins cause an increased risk of cancer? The evidence of this is spotty. The antioxidant effect of betacarotene seemed to decrease the incidence of cancer, but two studies involving lung cancer patients (randomized to receive betacarotene or not) were stopped early because of the increased incidence of cancer in the betacarotene group. This has never been done in breast cancer. It is important to keep in mind that oxidation is just the removal of electrons. If you take an electron away, you oxidize something. Vitamins are simply transport vessels of electrons; taking them from one place and bringing them to another. There are many actions of these compounds that we do not yet understand.

Is success judged by number of positive nodes found after neoadjuvant therapy? In the randomized study, the number of positive nodes found in the group of women who had therapy first, was lower. If you find a large number of positive nodes after neoadjuvant chemotherapy, patients tend to do worse.

How effective is tumor measurement after neoadjuvant therapy? It is true that the rate of tumor shrinkage and of growth is dependent on the tumor size. Larger tumors will tend to grow rapidly at first and then slow down. Measuring outcome (judging tumor size) by clinical exam is not as accurate as pathological measurement. MRI is helpful in some cases. If there is no residual cancer in the nodes and the breast after neoadjuvant chemotherapy, women with Stage III cancer only have about a 10% chance of recurrence. Chemotherapy may need to be given for a longer duration in some patients before it can be considered to have worked (or not).

Why aren't neoadjuvant patients given as much information as adjuvant patients regarding chemotherapy and follow-up care? Neoadjuvant patients need less preparation. They don't have to discuss surgery, so the process seems to move faster, with less time for the information given to sink in. This should not impact the amount of information given, or the time taken with the patient, but realistically, the number of patients seen by the average oncologist today is twice the number seen five years ago. This can amount to 20-30 patients seen daily. This is true for a variety of reasons: more people are coming in for therapies for cancers that used to be untreatable (Stage I breast cancer, lung cancer, advanced colon cancer). Managed care has required physicians to see more patients in order to keep revenue flat or rising. It is a major concern that although more people want to participate in their health care, they are unable to do so because there is no time for the physician to understand the patient personally. The bottom line is that new technology costs more in medicine as in other industries, and patients need to be willing to spend as much on their health care, if they want cutting edge therapies, as they are willing to spend more for computer technology.

What is the cause of weight gain in breast cancer patients? It is known that people do gain weight when they have breast cancer and are on chemotherapy. Tamoxifen might lead to weight gain in some people, but in the largest prevention trial to date there was no difference in weight gain between placebo and therapy groups. This is also true of the drugs Arimidex and letrozole. It is more likely that being ill tends to change ones lifestyle. Chemotherapy may cause a drop in activity due to depression and withdrawal. For many people tamoxifen therapy may also coincide with the onset of menopause.

What about clinical trials that are based on faulty premises? This has been the case with various observations about soy products on various populations. It may be that Japanese women who eat soy and don't suffer from hot flashes, never had hot flashes as symptoms of their menopause. It may be that soy by-products like pills and milk don't work in the same way as tofu does on symptoms of menopause. There is no substitute for observation, because even bad trials can tend to influence behavior for a long time. An example of this recently is the South African trials showing positive effects of bone marrow transplant. It was discovered that all of their data was falsified. The peer review process is usually thorough, but caveat emptor.

What is the trend in therapy for Stage I breast cancer? The trend seems to be to recommend therapy for smaller risk. However, a statistically significant result may not be clinically significant, e.g., is it worth taking chemotherapy for a 2% decrease in risk? The use of taxol in Stage II breast cancer has been shown to lower risk in patients with positive nodes. Now, out in the community, fully one half of Stage I patients are being given AC followed by Taxol. This is alarming and is not indicated in node negative patients.

What about cytokeratin cells in the bone marrow? Cytokeratin is a protein that is present in epithelial cells and breast cells, but not in bone marrow cells. If microscopic amounts of these cells are present in the marrow, your risk of recurrence is increased. Conversely, if you are node and cytokeratin negative your risk of recurrence is so low that you may need no further therapy. It is thought that bone marrow may serve as a reservoir for metastatic cells. Bisphosphonates, which prevent osteoporosis by strengthening the bone matrix, are being studied to see if they may protect against cancer cells lodging in the marrow, and consequently prevent recurrence of cancer. This in currently in clinical trials.

How do you get a large enough database to determine what works best for patients beyond your own institution? This is actually hard to do. We are looking all over the world, from Finland to U.S. to Canada to find a population that has the statistics necessary to do risk modelling. We need a robust database from which we can develop models of risk based on individual characteristics like tumor size, nodal status and risk of recurrence and how this changes with different types of therapy.