First Results from the International Breast Cancer Intervention Study (IBIS-I): A Randomized Prevention Trial

Abstract & Commentary

This is the first report from a randomized, placebo-controlled trial of tamoxifen for the chemoprevention of breast cancer in women deemed to be at high risk. In this study, 7152 women were randomized and 25% completed a full 5 years of treatment. Participants ranged in age from 35 to 70 years. The preponderance of participants was between the ages of 45 and 54 years. The overall odds ratio of breast cancer in the group treated with tamoxifen was 0.68 with a confidence interval of 0.50-0.92. There were 101 cases of breast cancer in the placebo group and 69 in the tamoxifen group. For those treated with tamoxifen, the odds ratio was 0.31 for ductal carcinoma in situ (5 in the tamoxifen group vs 16 in the placebo group; CI, 0.12-0.82), 0.75 for invasive breast cancer (64 vs 85 cases; CI, 0.54-1.04), 0.69 in those with estrogen receptor- (ER) positive breast cancer (16 vs 23 cases; CI, 0.47-1.02), and 1.00 in those with ER-negative breast cancer (19 cases in each arm).

With this apparent benefit came some risks. The rate of venous thromboembolism was 2.5 times higher in the tamoxifen-treated group (43 vs 17 cases; P = 0.001), with the rate of deep-vein thrombosis being 4.8 times higher (24 vs 5 cases; P = 0.0005). Gynecological and vasomotor symptoms were higher in the tamoxifen-treated group; P < 0.0001. There was a 2-fold excess of endometrial cancer in the tamoxifen group (11 vs 5; P = 0.2). Rates of pelvic ultrasonography, dilation and curettage, and hysteroscopy were significantly higher and there were more women with endometrial polyps in the tamoxifen group. Rates of hysterectomy were also higher in tamoxifen users, 4.2% vs 2.7%; P = 0.002. In premenopausal women, ovarian cysts and amenorrhea were more than twice as common on tamoxifen. Reports of monilial vaginitis were substantially increased in both premenopausal and postmenopausal women taking tamoxifen. A new finding was an increase in reports of brittle nails in the tamoxifen group. Inexplicably, the death rate from all causes was significantly higher in the tamoxifen group (25 vs 11; P = 0.03). Some of the increase in death appeared attributable to the increased rate of thromboembolism. The number of osteoporotic fractures was similar in both groups as were headaches and migraines (IBIS Investigators. Lancet. 2002;360:817-824).

Comment by Sarah L. Berga, MD

I don’t remember hearing about these study results on the news or reading about them on the front page of the newspapers. I guess that is the good news, because this is a difficult study to understand and the results are not likely to be easily reduced to a sound bite. However, if I had not gone back to scan the recent articles in the Lancet, I would have missed this intriguing data set. That would be bad news because this is the kind of information that rounds out our appreciation of the risk: benefit ratio associated with tamoxifen use in women who do not have breast cancer, but who are perceived to be at higher than normal risk for it in the future. Further, now that HRT has lost some of its glamour, there is a tendency to offer women other agents. In some cases, this includes tamoxifen as well as raloxifene. So it is important to know about its clinical effects.

Outlining the study results is one thing. It is easy to say that the data provide a sobering view of the risks and benefits of tamoxifen. A useful summary sound bite in this context is that the more tamoxifen is studied, the more the sense of peril replaces the promise. That is, in essence, the message that was formulated in the accompanying editorial (Kinsinger LS, Harris R. Chemoprevention of breast cancer: A promising idea with an uncertain future. Lancet. 2002;360:813-814). Chemoprevention of breast cancer is a good idea, but we are not yet ready to do it. In their opinion, there are 2 main barriers. The first is risk assessment strategies or tools. Who is a candidate? The second limitation is the agents available for chemoprophylaxis. Not addressed in this study or the editorial is whether chemoprophylaxis would need to be continued indefinitely. At what age would chemoprophylaxis start? Would the patient take it for 5 years, or more, or less? What would be the effect of any of these agents on the risks for other debilitating conditions such as osteoporosis, cardiovascular disease, or dementia? I guess the ultimate concern is that the use of a given agent would prevent breast cancer but hasten death. The data in this study are not reassuring. They suggest that chemoprophylaxis with tamoxifen might prevent the kind of breast cancers that are associated with a low chance of mortality while subjecting the users to a number of untoward health risks. No agent has been promoted as preventing aggressive breast cancers, that is, the ones that metastasize rapidly and kill.

Another of the other reasons I found these data of interest, of course, is that women who do have breast cancer also are frequently offered tamoxifen as adjunctive therapy. What might we tell them about the risks associated with its use? The data in this study provide some insight. Again, one worries that tamoxifen use prolongs survival from breast cancer but somehow increases the risk of dying while also diminishing quality of remaining life.

Ultimately, I am worried that we have started off on the wrong path, namely, that we have assumed that it is estrogen that is the cause of breast cancer. If this is not true, then we have made an egregious error in logic. The introduction of the study report starts with the following sentence: "The belief that oestrogen is the primary promotional factor for breast cancer has a long history and is now well established." Four citations are referenced at the end of this sentence. Three relate to the fact that the belief is old. One is a recent report of a prospective observational study of postmenopausal women. But then there are the data, including those of O’Meara (O’Meara ES, et al. Hormone replacement therapy after a diagnosis of breast cancer in relation to recurrence and mortality. J Natl Cancer Inst. 2001;93:754-762), that suggest that women with breast cancer who take estrogen preparations after the diagnosis of breast cancer survive longer (total mortality, 0.48; CI, 0.29-0.78) and have fewer recurrences (mortality, 0.50; CI, 0.3-0.85) and fewer contralateral breast cancers. What if tamoxifen’s effects upon breast cancer or the risk of breast cancer are due to its properties as a partial estrogen agonist? In the discussion, the IBIS group notes that "the agonist properties of tamoxifen account for its main side-effects of thromboembolic disease and endometrial cancer." Keep in mind that the study showed that tamoxifen prevented only ER-positive breast cancer. Thinking along these lines, I tried to determine whether the designation of a breast cancer as ER positive meant that it was positive for ERa. The article does not say, but I think that the assay for ER status is for ERa. The predominant ER in the endometrium is a and tamoxifen behaves as an agonist in that tissue. Would that mean that it serves as an agonist for breast cancers positive for ER? Not necessarily, but it would be fascinating if it did because it would suggest a role for ERa agents including traditional estrogens such as estradiol in the treatment or prevention of breast cancer. In short, this story is not ready for prime time. It is too soon to conclude that blocking estrogen action is a panacea for preventing breast cancer in part because it is too soon to say that estrogens cause breast cancer.

Dr. Berga is Professor and Director Division of Reproductive Endocrinology and Infertility, University of Pittsburgh.