WEDNESDAY, APRIL 8, 1198
"Tissue Research: Hopes and Concerns" ; "How to Read Your Path Report"

Historically, looking at tissue was a fairly perfunctory activity, just to confirm whether or not the patient actually had cancer. It is not clear as to how the tissue was used for decision making until the '50's, when scientists began noticing the value of lymph nodes in diagnosing cancer. In the '70's staging was accepted as a widespread way of categorizing cancer in general. Early staging techniques involved: size of tumor, its relation to other structures like the skin or muscle, whether lymph nodes were involved or whether metastasis had occurred. It was a fairly crude system, but one (for the most part) which is in use today, and seems to correlate well with overall outcomes. Many treatment decisions made today are based on TNM staging (size, location, nodes). Additionally, certain characteristics of the primary tumor seem to predict its "destiny": high tumor grade, whether it is invading blood vessels in the area, and many molecular markers, like HER2/neu. HER2/neu is the prototype of an amplified oncogene based on genetic amplification of material. This is a common hallmark of cancer cells which convey increased risk. There are many of these genetic markers, so analyzing tissue today has taken on a new level of importance.

Genetic (or genomic) instability causes mistakes in DNA during replication. It is probably one of the first things we are able to detect in a cancer cell microscopically. This causes the mutation rate to increase about a million fold. Enzymes repair DNA when these mistakes happen, but amplification and deletion (of tumor suppressors) tend to make the cell behave in a more malignant way. Cell abnormalities appear well before actual cancer is present. Scrapings from the cheek of smokers will show mutations in the P53 gene, an important tumor suppressor gene which is known to govern cell cycle and cell death. We have only scratched the surface in our understanding of these genes and how they work. To date we have identified 160 oncogenes and about 30 tumor suppressor genes, but there are many more. This is where the importance of studying tissue from patients comes in. There are multitudes of factors regarding how cancer behaves. In order to know how a particular marker affects outcome, we must study hundreds of patients. It took 30 large studies with hundreds of patients before people felt comfortable linking HER2/neu to outcome. No given marker can fully predict what a cancer cell will do, although there are some exceptions (like certain leukemias). In breast cancer no common molecular marker has been found. The value of studying human tissue and linking it to outcome has become clear to the research community. Most large universities have tissue banks, as does the NCI.

In the last few years a new technology has been developed that allows us to query the genome in an unprecedented way. You can now look at 10,000 genes at a time. Small pieces of DNA are placed on a glass wafer using robots. You can count how many genes are in the tumor. Huge amounts of information are now available on every tumor. The push to develop a large bank of tumor blocks brings with it the issue of patient confidentiality. Now that we can test for inherited genes, the information could be used against people. There is not currently enough legal protection against germ line genes like BRCA1 and 2. Patients must be provided with a thorough discussion on what will be done with their tissue. This should be included in their informed consent. Dr. Tripathy will soon be piloting a questionaire for patient input on a brochure for tissue research.

What is the meaning of tissue research for the individual versus advancement for the field in general? Few patients were aware that their tissue might have been used in some research protocol. There have been provisions in some past protocols where tissue could be pulled in an anonymous way to use without linking it to the patient, but we are moving away from collection without consent. The big problem with anonymous tissue is that it needs to be linked to patient outcome (early recurrence, etc.)

When research is done and you start to get results, but don't yet know what the relevance of the results are, how do you handle the information? Currently, the information will not flow back to the patient or physician. However if at a later time the information or marker is known to be important in decision making, it will be made available to the public and anyone can then go and have their tissue studied. This is considered to be the most appropriate method used today

Conflict comes from the fact that many patients have the desire to have all the information, even if they don't know what to do with it.

In California, state law requires that a block of tissue (average 2cm tumor removed in 3 parts, or 1/2 to 1 cm each = block)be kept for 7 years, some states are longer (New York requires 20 years). But some hospitals will not give patients their blocks after that time.

Grassroots efforts will be important in making changes in the way tissue is handled, and the number of patients who submit their tissue for research. Only a small amount of tissue is needed for most tests. Terminal patients are not able to take their tissue and do all tests possible. Just focused number of tests would be done to answer specific question. Studies of genetic markers are not as well publicized as clinical trials, but number of tests are huge. Most of these markers are not going to give information that would be useful to the patient today.

Today patients are not in control of their tissue. This could be changed if there was a coordinated grassroots effort for patients to initiate a tissue bank. Patients themselves would fill in the medical records and enter them in a database, which is currently very time consuming and consequently very costly. The next task is to prioritize the tests that would be done. This large scale centralized organization is very important to future research because currently a lot of information is being lost with small pockets of research being conducted all over the country but with no mechanism to share the information obtained. It would also be essential to identify the best institution to do the research, one which has the most expertise, facilities, etc. If you could get 1/10 of the women in the country to participate you would have 18,000 tissue samples per year!

The problem with studying a multitude of markers is that if you get one positive you are likely to overestimate its impact on the patient and her chance of recurrence, which will give a distorted sense of the benefit of therapy. If you are aware of that, then you should have the right to look at any markers you wish to. Exploratory analysis is sometimes called a "fishing expedition" because if you look at many markers and try to link them with outcome, you are likely to get some relationships purely by chance, not actually related to the event. Might be better to look for relationships with only a portion of your population and then use the remainder of the population as a validation set on several markers which seem to be promising.

In 80% of cases of early stage breast cancer, patients with rising CA15-3 did ultimately get metastasis. In 2-3% of the cases this will be a false positive. Early detection of this may not aid outcome. Therefore this test is often not done because it doesn't help in clinical decision making or improved outcome. More important is that we develop a therapy that can be applied as soon as a metastasis shows up.

There is now a standardized pathology report (Co-Path) which has been in place for about a year, although not all hospitals are yet on board. The characteristics that should always be included are:

• Tumor size
• Grade: Nuclear, Architecture (ability to form glands), Mitotic rate (number of cells in the act of dividing). SBR scale assigns 1-3 points for each feature, final score is sum of all 3
• Lymphatic vascular invasion
• Margins (Clear or not and size)
• Histopathologic type
• Number of Nodes (examined and number containing tumor)
• Largest Node
• Tumor outside the capsule
• Skin involvement
• ER and PR status

Why cells migrate to certain tissues is not clearly known, but is probably related to the genetic makeup of the cell. There are certain organs that breast cancer preferentially involves: lymph nodes, bone and lung, for example. Melanomas tend to involve the eye, probably mediated by specific proteins. Resistance to therapy is a question difficult to answer--researchers do not know how chemotherapy works. When you inhibit DNA, you send different signals to the cell, apoptosis being one signal. It may not be the chemotherapy itself that limits metastasis, as much as the cellular response to chemotherapy. Some people might not benefit because their cells are intrinsically resistant to start with; or resistance might have emerged upon treatment.

A good discussion about the terms used in the path report can be found in Dr. Ernie Rosenbaum's book, also Dr. Susan Love's Breast Book. The Community Breast Health Project in Palo Alto has weekly meetings where these things are often discussed. The group favored a classroom approach to learning about the path report before seeing the oncologist, so that the appointment can concentrate on treatment options.

Good News About Tamoxifen: It has been known for sometime that people on the tamoxifen adjuvant trials were getting about half the amount of new breast cancers in the other breast. About 6 years ago the tamoxifen prevention trial was started in which high risk women under 60 and all women over 60 were randomized with a placebo control group. 13,000 patients were enrolled. Early analysis showed a reduction of 45% in the number of cancers (150 in the control group vs. 80 in the tamoxifen group). The reduction was seen in women under the age of 50 as well as postmenopausal women. More endometrial cancers and thrombosis was found; the longer on the drug the higher the risk. See the NCI website for all the data: http://www.nci.nih.gov/ (Look for "What's New").

Roloxafene is another antiestrogen (developed as an anti-osteoporosis agent) without the uterine stimulatory effects. A large trial of 8000 postmenopausal women was done, looking at osteoporosis. The results indicated that the drug significantly improved osteoporosis, and unexpectedly, there were 80% less cancers in the treated group. These results are encouraging, but the study has only 2 years of followup.

A number of other compounds are poised for prevention trials. It is much harder to compare compounds in the adjuvant setting because metastatic recurrences can be fatal. Researchers are reluctant to take this step.