Raloxifene and Breast Cancer: The CORE Study

Abstract & Commentary

Synopsis: Raloxifene treatment of postmenopausal women with osteoporosis is associated with a lower incidence of estrogen receptor-positive invasive breast cancer.

Source: Martino S, et al. J Natl Cancer Inst. 2004;96: 1751-1761.

Martino and colleagues report the results of the Continuing Outcomes Relevant to Evista (CORE) trial, on the incidence of breast cancer. The MORE trial, the Multiple Outcomes of Raloxifene Evaluation trial, was a randomized, double-blind, multicenter clinical study of postmenopausal women with osteoporosis that reported a 72% reduction in estrogen receptor-positive invasive breast cancer in the treatment group after 4 years compared to placebo. The CORE study was designed to measure the impact of 4 additional years of raloxifene (60 mg/d), to begin during the fourth year of the MORE trial. Of the 7705 participants initially randomized in the MORE trial, 3510 women elected to continue raloxifene treatment (2336 completed the CORE trial) and 1703 continued on placebo (1106 completed the trial). During the 4-year CORE study, raloxifene treatment was associated with a 66% (HR = 0.34; CI = 0.18-0.66) reduction of estrogen receptor-positive invasive breast cancer in the treated group. There was no difference in estrogen receptor-negative tumors. Over the entire 8-year period, the reduction in estrogen receptor-positive cancers reached 76%. In the 8-year period, there was no difference in the number of deaths in the 2 groups.

Comment by Leon Speroff, MD

Overall, these results support a preventive effect of raloxifene treatment on the incidence of estrogen receptor-positive invasive breast cancer. The strength of this observation, however, can be questioned because of some problems within the study. One concern is the fact that the beginning of the CORE trial did not exactly coincide with the end of the MORE trial. From the end of the first 4-year study to the beginning of the next 4-year study, participants were not involved in study regulation for a time period that ranged from 2.6 to 62 months. During this interval, some participants experienced a long period without exposure to raloxifene, others used a standard regimen of hormone therapy, and those who experienced an adverse event, including breast cancer during this interval were excluded from the study. The characteristics of this interval and the impact on the results are issues essentially not addressed in this report.

The reduction in breast cancer observed in the 4 years of the MORE trial continued during the 4 years of the CORE study, and it is possible that the results in the second 4-year period reflect the effect of the initial 4 years. At the same time, the results are compatible with an ongoing impact beyond 4 years. Although the percentage of reduction is impressive, keep in mind that the actual numbers are not large: 21 cases of estrogen receptor-positive cancers in the placebo group and 15 in the treated group. In the entire 8-year period, the numbers added up to 58 in the placebo group and 40 in the treated group.

It is impossible to determine if there is a special high-risk group for whom this treatment is recommended. At the beginning of the CORE study, the women were assessed with the Gail breast cancer risk model. There was no difference between the treatment and placebo groups Gail predicted risk (about half were considered to be at high risk).

If medical treatment is truly preventive, one would expect to see a reduction in noninvasive breast cancer in the treated individuals. In the CORE trial there were only 9 cases (2 in the placebo group and 7 in the treated group) and in the 8-year period only 7 cases in the placebo group and 16 in the treated group, a number too small to allow confident analysis. On the other hand, an effect only on invasive disease without an effect on noninvasive disease keeps the possibility alive that raloxifene is affecting preexisting tumors. In contrast, a reduction in noninvasive disease has been reported with the preventive use of tamoxifen.1 One might conclude that the different results with tamoxifen and raloxifene reflect the risk level of disease in the 2 populations studied, high risk for breast cancer in the tamoxifen studies and low risk in the raloxifene osteoporosis study. However, as noted, assessment of breast cancer risk with the Gail model in the raloxifene study indicated that at least half of this population was also at high risk of breast cancer.

The evidence supports tamoxifen reduction of the risk for estrogen receptor-positive breast cancer, but at the same time, tamoxifen should not be recommended as a preventive agent, except for women at very high risk. This conclusion is based upon the degree of reduction in risk compared with the incidence of side effects. The evaluation by the National Cancer Institute is very helpful.2,3 Because the risks associated with tamoxifen (endometrial cancer, stroke, pulmonary embolism, and deep vein thromboembolism) increase with age, balancing the risks and benefit indicates that tamoxifen is best for younger women with an elevated risk of breast cancer (an increased relative risk of approximately 1.7). A similar conclusion was reached by a working group of the American Society of Clinical Oncology.4 This means that only a relatively small number of women qualify, about 5% of American white women and 0.6% of black women.3

The data are too limited to support the use of raloxifene as prophylactic treatment, and a stronger position awaits the outcome of the STAR trial comparing tamoxifen with raloxifene and the RUTH trial assessing the effect of raloxifene on both cardiovascular disease and breast cancer. The Medical Research Council of the United Kingdom and the National Cancer Institute of the United States have reached similar conclusions.

It is well-recognized that raloxifene shares with tamoxifen and estrogen about a two-fold increase in venous thrombosis. However, the increase observed in the 8 years of the combined MORE and CORE population did not achieve statistical significance, a problem of relatively small numbers because of the infrequency of this event. It is noteworthy that the 9 cases of pulmonary embolism all occurred in the raloxifene group.


1. IBIS INVESTIGATORS. Lancet. 2002;360:817-824.

2. Gail M, et al. J Natl Cancer Inst. 1999;91:1829-1846.

3. Freedman A, et al. J Natl Cancer Inst. 2003;95: 526-532.

4. Chlebowski R, et al. J Clin Oncol. 2002;20:3328-3343.

Leon Speroff, MD, Professor of Obstetrics and Gynecology, Oregon Health Sciences University, Portland, is Editor for OB/GYN Clinical Alert.