Helen Schiff, A NYC Advocate

Speaks On Tamoxifen in Healthy Women

I would like to dedicate this talk to Lois Kasper, My dear friend from SHARE who just died May 18th. Even when she had metastatic disease she helped other women. She took hot line calls from women with advanced breast cancer; she appeared on the Charlie Rose Show along with Fran Visco, Dennis Slamon and Larry Norton; And she allowed a local TV station follow her life, the life of a young women dying of breast cancer. I will never forget her.

When I first read in the newspaper that the P-1 trial had been halted early because tamoxifen had reduced the incidence of breast cancer by half I was very excited. Excited by the new possibility this drug offered. Tamoxifen is clearly an advance in the treatment of breast cancer. Now it has been proven to be a chemopreventative as well.

But as I began to look at the results some red flags went up. There were three deaths from pulmonary embolism in the tamoxifen arm and none in the placebo arm. The incidence of pulmonary embolism was 18 in the tamoxifen arm versus six in the placebo arm. Women also had an increased risk of deep vein thrombosis, endometrial cancer and stroke. I asked myself the question: Was this drug cutting the incidence of one disease but creating several more? If tamoxifen has life threatening side effects, shouldn't mortality have been the trial's endpoint, not just breast cancer incidence?

Then I read that there was a median follow-up of only four years. I wondered if that was really long enough to test a drug for prevention. I began to question whether we really had enough data to give this drug to healthy women, the overwhelming majority of whom would never get breast cancer, yet they would be subjected to tamoxifen's toxic and non toxic side effects. I began to think of the results of the P-1 trial as breakthrough on a scientific level: a proof of principle that chemoprevention does work. But is it ready to be given to women?

What I found extremely scary was the oft repeated quote in the media stating that 29 million women were candidates for tamoxifen. This astounding statistic generated by Zeneca, the manufacturer of tamoxifen, was repeated by the NCI, the NSABP and the FDA. 29 MILLION WOMEN is almost 40% of the female population between the age of 35 and 89 in the U.S.

Things began to move very fast after the P-1 trial was stopped. Too fast, in my opinion. Before the P-1 results were even peer reviewed and published, the FDA called a public meeting of ODAC, its independent advisory committee. Advocates, including myself, voiced their concerns at the meeting. These were real world concerns.

1) Are we talking about prevention or are we talking about risk reduction.

2) Do we know enough about this drug to give free rein to Zeneca, to advertise it directly to such a large group of women? (Many of whom greatly women overestimate their risk of breast cancer. )

3) Do primary care physicians and gynecologists have the expertise to prescribe this drug? (After all they will be the ones, not oncologists, whom women will consult.)

4) We voiced our concerns about women with BRCA 1 or 2 mutations. Here certainly, I thought, was a group at high enough risk for tamoxifen. But there were no data available on tamoxifen's effect on mutations carriers. We will still not have that data for at least another year.

Five out of the 6 breast cancer advocates who spoke at the ODAC meeting opposed approval of tamoxifen for this new indication.

Dr. Trever Powles, who heads the British tamoxifen prevention trial, addressed the meeting. His trial was much smaller but lasted 2 years longer than the P-1. It failed to find any benefit for tamoxifen. I was struck by what he said: was it worth giving 6,681 women tamoxifen to prevent 86 breast cancers? Was it worth it, ESPECIALLY when tamoxifen's effect appeared to be centered on SMALL estrogen receptor positive tumors with NEGATIVE lymph nodes, the most EASILY CURABLE type of breast cancer.

I was also impressed with the point raised by Dr. Richard Simon of the NCI. He asked whether the eligibility criteria of the P-1 trial, which was a five year breast cancer risk of 1.7% for women between the ages of 35 59 and for all women over the age of 60, he asked was this really an appropriate high risk threshold, since the women who had actually enrolled in the trial had an average risk that was 3.2%, TWICE as great.

At the end of the meeting Dr. Leslie Ford of the NCI announced the existence of a computer disk which women and doctors could slip in their computer, take a quick test and find who is at high enough risk to take tamoxifen.

The FDA approved tamoxifen a month later as a drug that will reduce the risk of--not prevent-- breast cancer. There was also a change in the original eligibility criterior from all women over 60 to women over 60 with an average or higher risk of breast cancer.

Usually the manufacturer is the driving force behind a drug's approval by the FDA. But, in this case, the NCI, the NSABP and the FDA appeared to be actively joining, in what I saw as an imprudent and dangerous rush, to approval of tamoxifen for this new indication.

A year after the P-1 was halted, the American Society of Clinical Oncologists became the first physician group to conduct a critical assessment of its results. The tamoxifen working group of ASCO, of which I was a member, found that though tamoxifen did reduce the incidence of breast cancer, there was no STRONG or CREDIBLE EVIDENCE that it would reduce breast cancer mortality. Furthermore, there was no STRONG or CREDIBILE EVIDENCE of an overall health benefit. This point is important because serms are supposed to have hear and bone benefits. There was no heart benefit and the increase in bone density was small- not nearly as beneficial as Fosamax. Unfortunately, although it did question the 1.7% high risk threshold it did not unequivocally state that it was too low. For this reason I withdrew my name from the assessment.

About a year and a half after the P-1 was halted, I was dismayed to read in the Journal of the National Cancer Institute about a new benefit/risk assessment to determine who should take tamoxifen as a chemopreventive. By now, the Zeneca nationwide AD campaign, directed at women and doctors had been in full swing for almost a year. The 1998 media hoopla surrounding the P-1 had left the impression that all women over age 60 were candidates for this drug. Now comes the careful risk/ benefit assessment that SHOULD have preceded the FDA approval.

I was outraged to learn that this assessment had actually begun before ODAC met to make its recommendation to the FDA approval. Wouldnt it have been wise to wait for the results before approving the drug? The findings of the new assessment were astounding: candidates for tamoxifen were a far far smaller group than we had originally been led to believe. Almost all women over the age of 60, with a uterus, were now excluded because the odds of getting tamoxifen-induced endometrial cancer is equal to the number of breast cancers prevented. And the stroke incidence, as well as other vascular events, among older afro-american women is so high, as to make tamoxifen too risky for almost all afro americans over age 60.

As it turned out women who had volunteered for the P-1 trial experienced less of tamoxifens life threatening side effects than the average american black or white women. This reflects a phenomena know the healthy women effect. People who volunteer for these kinds of trials are generally in better health. The p-1 data does show that a small group of VERY VERY HEALTHY women women over sixty WITH a uterus might benefit from tamoxifen if their risk of breast cancer was extraordinarily high, that is a 5% to 7% five year increased risk

This new assessment showed that a major revision was in order for the risk disk. But is has yet to be revised. 100,000 copies having already been distributed.

I question the assumption in the risk/ benefit analysis that tamoxifen is safe for premenopausal women. The trial lasted only 4 years. Remember it took 20 years to discover that tamoxifen caused a 4 fold increase in endometrial cancer in postmenopausal women. The ten million women years of tamoxifen use in women with breast cancer, which is often cited as evidence of its long term safety profile, was almost exclusively in postmenopausal women and women with metastatic disease who usually dont live longer than a few years. There is hardly any longterm data for premenopausal women.

PAUSE Any decision to take long-term drug therapy requires a thoughtful consideration of risk-to-benefit ratio, but in this case, even more care must be taken because the target population is healthy. Fast tracked approval, tamoxifen was fast tracked by the FDA, might make sense for some drugs needed by people who are dying, but much slower and more deliberate process was in order for a drug that is meant to be given to thousands if not millions of healthy women.

PAUSE The manner in which clinical trial results are conveyed to the public, becomes even more crucial. When a controlled trial shows that a drug is superior to a placebo, the accepted method of conveying successful results is in relative terms. There were nearly 50% fewer cases of breast cancer in the tamoxifen group. This is translated to doctors and woman as: you can cut your risk of getting breast cancer in half. But in the interest of informed consent, women should have this information converted for them in absolute terms: The P-1 trial found the rate of breast cancer in the placebo arm was 2.6%, as compared to 1.3% in the tamoxifen arm so the absolute risk reduction was only 1.3%. For women at a 1.7% risk, the so called high risk threshold, the absolute risk reduction would be even lower- .85%

Youll see, in the following ads, how Zeneca blurs the difference between relative and absolute to sell women not only the fear of breast cancer, but also to exaggerate tamoxifens safety.

SLIDE 1: Here is a young women in a bra with a headline that says: If you care about breast cancer, care more about being a 1.7 than a 36B. Know your breast cancer risk assessment number. The results will give you a number that estimates your chances of developing breast cancer over the next five years. The ad tells you that a score of 1.7 or above is considered high risk. NO WHERE DOES IT SAY what this number or score means. If Zeneca actually told women that the 1.7 means: YOU HAVE A 1.7% CHANCE OF GETTING BREAST CANCER in the next five years women might not be so keen about going on a five-year drug regimen. Why take this drug when you have a 98.3% chance of NOT getting breast cancer. By the way ads in medical journals do say what the 1.7 is. I filed a complaint with the FDA about the misleading use of the word score when Zeneca really means a percentage of your risk. I just received a letter from the FDA, saying that my complaint has merit.

SHOW SLIDE 2: This one I received in the mail. It says: Do you try not to think about it? Do you forget about it. Do you never forget about it? Do you think positive? Or stick your head in the sand? Do you distract yourself with a good novel? A movie? A chocolate bar? Do you dream about it? Despair about it? Rage about it? Do you dread it? Is this what you do about breast cancer? Or do you do something about it?

SLIDE 3: Here is part of the text from this ad. You will see Zenecas misleading shift from relative to absolute risk. The ad says: Nolvadex (which is the brand name for tamoxifen) was proven to help reduce the chance of a first occurrence of breast cancer by 44%. This is a relative figure.

SLIDE 4: Juxtaposed to this, it says Nolvadex isn't for every woman at high risk. In the study, women taking Nolvadex were 2 to 3 times more likely to develop endometrial cancer or blood clots in the lung or legs, although these occurred in less than 1% of women. This is the absolute figure.

Most women would figure I cut my risk of breast cancer by 44% and I have only a less than 1% risk of life-threatening side effects. That looks pretty good. But it only looks good because they are mixing the relative and absolute numbers.

The National Women's Health Network filled a complaint with the FDA against this misuse of statistics. The FDA has responded that this complaint also has merit.

Obviously, ads like these are more about selling drugs than helping people make informed decisions. Zeneca, obviously, is more concerned about expanding tamoxifen's market than women's health. Unfortunately, it takes about 6 months to retract an inaccurate ad.

I take issue with the underlying assumptions of the P-1 investigators. They have left out the woman's point of view. In their published results the investigators state that the diseases tamoxifen causes, such as endometrial cancer and pulmonary embolism, are not as bad as breast cancer. They cavalierly suggest, for example, that endometrial cancer is simply dispatched with a hysterectomy. Yet this is major surgery with a small rate of mortality and a large rate of complications, including sexual dysfunction, bladder and bowel incontinence.

Quality of life is an even more important issue here. The woman with breast cancer might tolerate terrible tamoxifen-induced hot flashes, suppression of libido, and vaginal discharge. But do we want to inflict that on a large number of healthy women?

Now healthy women are being recruited for the STAR trial. Tamoxifen will be compared against raloxifene. I worry that the results will not be meaningful because there is no placebo arm in this trial. There will be no longterm followup from the P-1 trial either because when it was halted back in 1998, the participants on placebo were actively encouraged to join the STAR trial.

We need a better model to determine who is really at high enough risk to take a drug with serious side effects. Right now we have a situation where only a very samll number of women who qualify to take tamoxifen will ever get breast cancer and of the women who get breast cancer most, between 70-80% would not be considered at high enough risk to take tamoxifen.

Women with DCIS and LCIS are an examples of very high risk women for whom tamoxifen maybe appropriate, yet even for them the risks and benefits must be carefully weighed. WHAT WE REALLY NEED is non toxic chemopreventative agents and lifestyle changes that will reduce womens breast cancer risk. That means research into the causes of bc. Both er+ and er-. Does all breast cancer start out as er+ or is their pathogenesis different?

AND for immediate action, we need, at the very least, to get the NCI to amend the risk disk. (I just heard the day before yesterday that a new disk will be out in a few weeks.) Even more important, we need to pressure the FDA to change the tamoxifen label to reflect the latest risk/ benefit assessment. The label says nothing about the questionable safety of this drug for older women. We need to keep monitoring Zenecas ads.

Advocates want new breast cancer treatments. We want new and safe chemopreventative agents. But more important than getting them fast is getting it right.

Questions and Answers on Tamoxifen

For newly diagnosed women by Helen Schiff, Advocate

Remember we are NOT Doctors and have NO medical training.

This site is like an Encyclopedia - there are many pages, many links on many topics.

Support our work with any size DONATION - see left side of any page - for how to donate. You can help raise awareness of CAM.