Questions and Answers about the Overview and Tamoxifen

The aim of this large study was to assess reliably the effects of tamoxifen on cancer and, particularly, on long-term survival in women who have just had apparently successful surgical removal of a breast cancer. (Tamoxifen could protect such women against any undetected fragments of the breast cancer that had already spread elsewhere). This aim is achieved because the study is so large and thorough.

The only side-effects shown are those that involve cancer or particular causes of death: since such side effects are rare, they could be assessed reliably only in a large study. This study is of tamoxifen for women who have already had breast cancer: thus, it differs from recent reports of breast cancer prevention by tamoxifen in previously healthy women.

A graphical summary of the study results is given at the end of this document.

Q. Is this the largest randomised study of cancer treatment ever undertaken?

A. Yes. The Early Breast Cancer Trialists' Collaborative Group (EBCTCG) currently has data on 130,000 women who had breast cancer. The present report and press release relate to the 37,000 in trials assessing tamoxifen, of whom about 30,000 had hormone-sensitive breast cancers.

Q. How many countries were involved in this assessment of women with breast cancer, how many trials, how many women and how much has it cost?

A. 15 countries, 55 trials, 37,000 women (of whom about 30,000 had hormone- sensitive cancers). The study cost of about £1m over the past 15 years has been paid for by the Imperial Cancer Research Fund.

Q. What was the initial objective of this study 15 years ago, and has this changed?

A. The main aim in 1983-85, when these worldwide "meta-analyses" (or overviews) first began, was to see if tamoxifen affected survival at all. Most people thought it didn't. Now that we know it does, we are studying the effects of prolonged tamoxifen (around 5 years) on long-term survival (more than 10 years), and finding excellent results in a wide range of women who had breast cancer.

Q. What is a "meta-analysis"?

A. Results from all randomised trials that have addressed the same question are combined. This provides very reliable evidence since it is based on much larger numbers than are in any one of the contributing trials.

Q. What are the three most important findings?

A.

1. Five years of tamoxifen substantially improves 10-year survival.
2. A wide range of patients benefit substantially: not just older patients, but also the young (i.e. premenopausal); and, not just those with disease originally found only in the breast, but also those with breast cancer that has spread to the local lymph glands. Tamoxifen also provides added benefit irrespective of whether or not chemotherapy has been used.
3. Tamoxifen halves the incidence of new cancers in the opposite breast. The EBCTCG report in 1992 was the first in the world to show prevention, and the present results consolidate the finding.

Q. Why was it thought that tamoxifen was not effective in younger women? What was the logic behind that?

A. The studies in young women were not big enough and tamoxifen was not given for long enough for evidence of benefit to emerge. This apparent lack of effect, which was incorrect, was then rationalised by suggesting that high premenopausal hormone levels might overwhelm the drug.

Q. If most doctors really are not prescribing tamoxifen to young women very much, how did you accumulate the evidence to prove that it works for them?

A. Because a few thousand younger women did agree to participate in research trials out of the millions of young women with breast cancer over the past 25 years.

Q. If doctors thought tamoxifen was not effective in younger women, why have they been giving it to 35-year-olds in primary prevention trials?

A. To get evidence: doctors can think what they like, but until they have large-scale randomised evidence, they may well be wrong.

Q. Have any of the findings in 1992 been modified as a result of a further 5 years' follow-up?

A. Yes: there is now much stronger evidence that 5 years of tamoxifen is better than 1-2 years. There is now a wider range of patients with definite benefit and there is even better evidence that the life-threatening side effects are much smaller than the improvement in survival.

Q. How widely is tamoxifen currently used in: (a) developed countries? (b) developing countries?

A.(a) About a million women use it. That's not enough: young women in particular seem to be under-treated.

A.(b) Its use is sometimes limited by price, even where generic tamoxifen is available, but tamoxifen is increasingly widely used in China, India and South America.

Q. Where it is not being used, what are the main reasons?

A.

1. The cost, in poor countries.
2. Doctors don't realise how effective it is: in women aged over 50 it has a far bigger benefit than chemotherapy (although these are not mutually alternative treatments).
3. Failure to use it in young women, in those without spread to lymph nodes and in women who have had chemotherapy. The present results show that all these groups can benefit from tamoxifen. Widespread dissemination of this new evidence should help ensure that they are treated.

Q. What proportion of all breast cancers occur in premenopausal women, and is it the same in developed and developing countries?

A. About one fifth in developed countries and about a half in developing countries. The difference is largely due to differences in the age distribution of these different populations.

Q. What proportion of premenopausal women with breast cancer would you expect to benefit from tamoxifen i.e. are hormone-sensitive?

A. About half.

Q. What exactly is meant by hormone-sensitive?

A. If the cells that make up the breast cancer are "sensitive" to female hormones, the growth of these cells will be stimulated by these hormones. This is true for most women with breast cancer. In order to be sure whether or not a particular woman has hormone-sensitive breast cancer, some of her tumour cells would be examined in a laboratory.

Q. Should all premenopausal women who have hormone-sensitive breast cancers (or where the hormone receptor status of the original tumour was not measured) receive tamoxifen?

A. We don't say "should". Each woman's treatment is an individual decision for her and her doctor: But, if such a woman does receive tamoxifen then her risk of death from breast cancer over the next 10 years will be reduced, on average, by at least a quarter.

Q. Should tamoxifen be given in addition to chemotherapy or in place of it for breast cancer? If it is given in addition, would that incur extra benefit?

A. Again, we don't say "should", but tamoxifen reduces whatever breast cancer death risk remains after chemotherapy. Doctors who want to give chemotherapy without tamoxifen can always start tamoxifen after the chemotherapy has ended.

Q. Is it appropriate for women who have had ovarian ablation (removal or suppression of the ovaries) for breast cancer to be given tamoxifen?

A. As tamoxifen helps post-menopausal women, presumably it also reduces whatever breast cancer death risk remains after ablation.

Q. Does it matter what type of surgery a woman had to remove her breast cancer?

A. No, tamoxifen is beneficial whether a woman had a mastectomy to remove her whole breast or a lumpectomy to remove just the tumour.

Q. Does it matter whether a woman has had radiotherapy after her surgery?

A. No, tamoxifen is beneficial whether or not a woman receives radiotherapy.

Q. In the 1992 overview, you stated that 1,000 extra lives a year and 10,000 worldwide could be saved if women with early breast cancer had appropriate cell- killing or hormonal treatment as well as surgery. How much could this figure be increased if doctors take note of your latest results and prescribe accordingly?

A. In 1992 we actually under-estimated the benefits. Tamoxifen alone is probably already saving 20,000 lives a year. The present results suggest that the numbers of lives saved could be doubled if tamoxifen was much more widely used.

Q. What side effects of tamoxifen might affect cancer or survival?

A. It increases endometrial cancer (cancer of the lining of the womb) and pulmonary embolus (blood clot in the lungs), but it decreases the chances of getting a new breast cancer in the opposite breast. In women who have had breast cancer, the decrease in the number of new breast cancers is bigger than the increase in endometrial cancer.

Q. In terms of 10-year survival, what is the balance of risks and benefits?

A. An additional 3 women per thousand die from endometrial cancer or pulmonary embolus. In women with hormone-sensitive cancer 5 years of tamoxifen prevents about 30 times as many deaths as it causes. In women with untested breast cancers, it prevents about 20 times as many. (Even in women with no hormone receptors on their breast cancer, it causes fewer new cancers than it prevents, but these increases and decreases involve relatively small risks and benefits).

Q. If doctors act on your evidence, more women will be given tamoxifen. Do you plan any quality of life studies?

A. No. These questions need to be studied carefully in small populations. Our study investigates cancer incidence and causes of death only, which need to be studied in large populations.

Q. Are the side effects different in premenopausal women and postmenopausal women?

A. Interference with normal menstrual cycle can cause symptoms, but is not something that we have useful data on. Obviously, if a woman is planning to become pregnant, she should discuss her situation carefully with her doctor.

Q. Do the side effects of tamoxifen go away once treatment is stopped or does an increased risk (e.g. of endometrial cancer or thrombosis) remain?

A. The numbers are so small it is difficult to know: these risks may well diminish, but we'll have better evidence from the next 5-yearly follow-up of these trials.

Q. There were concerns about a link with liver cancer and bowel cancer. Has the overview ruled these risks out?

A. We have shown there is no good evidence that either risk is real.

Q. What would you say, with present knowledge, is the optimal length of treatment and is this the same for older and younger women?

A. In terms of recurrence, 5 years is definitely better than just 1 or 2 years and that seems increasingly likely to be true for survival as well (although we won't be absolutely sure until the next follow-up of these trials in about the year 2,000). Unfortunately, however, we won't know until 2010 whether or not 10 years of treatment is better than 5.

Q. Are there lower or upper age limits for giving tamoxifen?

A. No, not if a woman has already developed a hormone-sensitive breast cancer.

Q. Are there any breast cancer patients for whom tamoxifen doesn't work?

A. Tamoxifen reduces the risk among women with breast cancer that is hormone- sensitive (or in which hormone receptor measurements have not been made). But, if their risk of breast cancer recurrence and death is small (as, for example, with a small cancer detected by breast screening) then so too would be their absolute benefit. (Even women whose original breast cancer had no hormone receptors at all can, however, still be protected against getting a new cancer in the opposite breast.)

Q. Does tamoxifen prevent recurrence or only delay it?

A. Our 10-year results strongly suggest prevention: but in 10 years' time the 20- year results will provide a clearer answer.

Q. You say tamoxifen will improve 10-year survival even for women whose hormone receptor status is unknown. Does this apply both to younger and to older women?

A. Yes. Half the younger women and three-quarters of the older women will, in any case have hormone-sensitive cancer.

Q. Even if tamoxifen is not of much benefit in women whose original tumour was proven to have no hormone receptors at all, might the gain be worthwhile by reducing the risk of a new cancer in the other breast?

A. It might be, but the absolute benefit would be considerably smaller than in other women with breast cancer.

Q. Is a woman whose original cancer had no hormone receptors still at risk of developing a hormone-sensitive cancer in the other breast?

A. Yes.

Q. You say (in the press release) that as most breast cancers are hormone-sensitive, tamoxifen can still be used even if no hormone tests were done. But some of those women on whom no tests were done would in fact have no hormone receptors at all on their original cancer. Then you would be giving a drug which probably gives little or no protection against their original cancer, but which can have side effects such as endometrial cancer and blood clots on the lungs. Is that justifiable?

A. Even if the original cancer was cured, or was not hormone-sensitive, tamoxifen would prevent twice as many new breast cancers as it would cause endometrial cancers and, overall, it would probably do at least as much good as harm in terms of survival. But the big gain would be that the many women whose cancers are hormone- sensitive would be protected.

Q. If a woman reading these reports has been treated for hormone-sensitive breast cancer and not been offered tamoxifen would you advise her to go back to her consultant and discuss it?

A. Yes, if the risk of recurrence is still substantial. Indirect evidence shows that tamoxifen can approximately halve whatever risk remains.

Q. Is HRT a reason not to give tamoxifen (or is the use of tamoxifen a reason not to give HRT)?

A. As far as we know, this question hasn't been properly answered.

Q. If a woman, reading these reports, had taken tamoxifen for one or two years for hormone-sensitive breast cancer but the drug had then been stopped for reasons other than side effects, would you advise her to go back to her consultant to discuss it?

A. Yes, if the risk of recurrence is still substantial, since indirect evidence from these trials show that tamoxifen can approximately halve whatever risk remains.

Q. Does tamoxifen have to be given reasonably quickly after surgery to be beneficial?

A. It is best started soon after surgery (or, at least, after any chemotherapy that may be given). But there may well be benefit from later use if the risk of recurrence is still substantial. See above.

Q. How much has tamoxifen contributed to the fall in breast cancer deaths in those countries where death rates have fallen?

A. In the UK, for example, probably more than half of the sudden fall in breast cancer death rates during the 1990s is due to tamoxifen. If so, then tamoxifen is already preventing one or two thousand UK breast cancer deaths a year.

Q. What is the present state of studies on giving tamoxifen for longer than 5 years, and when will results be available?

A. There were three old trials that were much too small to be reliably informative, plus two large trials - ATLAS and ATTOM - that are still recruiting and won't give reliable results until 2010.

Q. Is it reasonable for doctors to give tamoxifen for longer than 5 years until there is a reliable answer?

A. If they don't know whether to stop or continue then they have to guess: either guess is reasonable - but better still is to invite the women to join a trial addressing this question.

Q. Many cancer drugs fail because a patient develops resistance to them after a while: if a woman with node-negative breast cancer and, therefore, a low risk of recurrence, is given tamoxifen but the cancer recurs, might she be worse off overall because of drug resistance?

A. Not on average. These trials show that the best policy on average is to start treatment early.

Q. How much does it cost to treat a woman with tamoxifen?

A. The prices for one 20mg tamoxifen tablet per day vary within one country and between countries. In Britain 5 years of tamoxifen costs 200-400 pounds (300 pounds is about 500 dollars), and in most other countries it costs about 200-800 US dollars. But, in Japan 5 years of tamoxifen costs about 3,000 US dollars (400,000 yen) and in the United States it costs about 6,000 US dollars.

Q. How can you estimate the drug cost per life saved?

A. Since 5 years of treatment produces a survival advantage of 1 in 12, we just multiply its cost by 12.

Q. Can the drug be afforded, even by developing countries?

A. Yes, in almost all developing countries a proportion of the breast cancer patients can afford tamoxifen, or their medical services can afford to give it. But, in many developing countries there are women with breast cancer who need tamoxifen but can't afford it, and don't get it.

Q. How is tamoxifen used?

A. Tamoxifen is usually taken once a day as a tablet.

Q. Does the dose of tamoxifen matter?

A. Most of the women in this study took either 20mg or 40mg of tamoxifen a day. The benefits of tamoxifen seem to be about as large for both of these doses, and 20mg a day is now commonly prescribed. What does make a difference, however, is how many years the tamoxifen is taken for.

Q. How does tamoxifen work?

A. There may also be other ways in which tamoxifen works, but one of the most important is that it can attach itself to the hormone receptors in any remaining breast cancer cells and get in the way of hormones which might make those cells grow out of control.

Q. Will the overview continue and report at 20 years follow-up?

A. Yes - and at 30 years, if we're still alive ourselves!

Q. What specifically will you be looking for at the 20 year point? Are there new questions to be addressed?

A. We need to know, with respect to long-term survival with tamoxifen, what happens after year 10 to the main effects and to any side effects. Also, results will start to emerge from direct comparisons of different durations of tamoxifen treatments.

Q. Where can one find out more about this study?

A. The full report of this study is published in the Lancet medical journal (16 May 1998 issue) and will also be available to those who subscribe to the Lancet website. Also, there will be some details available on the Oxford CTSU website.

Q. Are there organisations with whom women can discuss tamoxifen?

A. Yes.

Results Summary

Below is a graphical summary of some of the numerical results from the study.

Avoiding RECURRENCE during the first decade
All women with potentially hormone-sensitive breast cancer
% still without breast cancer recurrence		~5 years of TAMOXIFEN CONTROL (no tamoxifen unless disease recurs)
		7400 women: 16% difference in breast cancer recurrence (with SD 1.2) P < 0.00001
	years

Avoiding DEATH during the first decade
All women with potentially hormone-sensitive breast cancer
% still alive		~5 years of TAMOXIFEN CONTROL (no tamoxifen unless disease recurs)
		7400 women: 8% difference in breast cancer mortality (with SD 1.1) P < 0.00001
	years

Avoiding RECURRENCE during the first decade
Age < 50 at surgery
% still without breast cancer recurrence		~5 years of TAMOXIFEN CONTROL (no tamoxifen unless disease recurs)
		1300 women P < 0.00001
	years

Avoiding RECURRENCE during the first decade
Age 50+ at surgery
% still without breast cancer recurrence		~5 years of TAMOXIFEN CONTROL (no tamoxifen unless disease recurs)
		6100 women P < 0.00001
	years

Avoiding RECURRENCE during the first decade
Cancer originally detected only in breast
% still without breast cancer recurrence		~5 years of TAMOXIFEN CONTROL (no tamoxifen unless disease recurs)
		5200 women with "node-negative" disease P < 0.00001
	years

Avoiding RECURRENCE during the first decade
Cancer in breast and local lymph glands
% still without breast cancer recurrence		~5 years of TAMOXIFEN CONTROL (no tamoxifen unless disease recurs)
		2200 women with "node-positive" disease P < 0.00001
	years

Avoiding RECURRENCE during the first decade
Tamoxifen vs. not, without any chemotherapy
% still without breast cancer recurrence		~5 years of TAMOXIFEN (without chemotherapy) CONTROL (no tamoxifen or chemotherapy unless disease recurs)
		6500 women P < 0.00001
	years

Avoiding RECURRENCE during the first decade
Tamoxifen + chemotherapy vs. same chemotherapy alone
% still without breast cancer recurrence		Chemotherapy + ~5 years of TAMOXIFEN Same chemotherapy (but no tamoxifen unless disease recurs)
		900 women P < 0.00001
	years

Breast Cancer Mortality 1950-95/96
(mean of rates at ages 50-69)
Annual death rate per 100,000 women		UK USA
	year

• Substantial protection against recurrence of original cancer, and against death.

• Decrease in chances of new cancer in the other breast ("contralateral") is bigger than increase in chances of new cancer in the womb ("endometrium").

• No apparent effect on other cancers, or on aggregate of all other causes of death.