Clinical Alert: Adjuvant Therapy of Breast Cancer - Tamoxifen Update

National Cancer Institute (NCI)
November 30, 1995

Abstract:

The National Cancer Institute wants to bring new information regarding the optimal duration of adjuvant tamoxifen therapy to the attention of clinicians. Recent results from the National Surgical Adjuvant Breast and Bowel Project (NSABP) Protocol B-14, that evaluated 5 years versus 10 years of adjuvant tamoxifen for early stage breast cancer, indicate no advantage for continuation of tamoxifen beyond 5 years in women with node-negative, estrogen receptor-positive breast cancers. In view of the proven benefits of 5 years of adjuvant tamoxifen, this treatment should continue to be administered whenever appropriate to women with early stage breast cancer. However, the new data suggest that more than 5 years of adjuvant treatment is not warranted in routine clinical practice in this patient population.

The full-text of the announcement is available via the National Cancer Institute's fax, electronic mail, World Wide Web, and PDQ services. To receive the announcement by fax, use NCI's CancerFax by calling 301-402-5874 from your fax machine telephone and, following the voice prompts, enter code 400122 when asked. To receive the announcement by e-mail, use NCI's CancerNet by sending an Internet electronic mail message to cancernet@icicc.nci.nih.gov with the following in the BODY of the message: cn-400122. The document will be returned in electronic mail. To access the announcement via World Wide Web, use WWW browsing software to point to the URL //wwwicic.nci.nih.gov/ and look for late-breaking information. Health professionals may contact the PDQ Search Service at 1-800-345-3300 and the general public may contact the Cancer Information Service at 1-800-422-6237 to have information from PDQ sent to them.

The full text of this alert has been mailed to all libraries that are members of the National Network of Libraries of Medicine.

Full Text:

The following is the text of a clinical announcement released by the National Cancer Institute on November 30, 1995.

The adjuvant administration of tamoxifen has significantly improved disease free survival and overall survival in women with early stage breast cancer. In 1981, the National Surgical Adjuvant Breast and Bowel Project (NSABP) initiated Protocol B-14 in order to determine the efficacy of tamoxifen in women with primary breast cancer with histologically negative nodes and estrogen receptor- positive tumors. Two thousand eight-hundred and ninety-two patients were randomized to receive either tamoxifen or placebo. The results initially published in 1989 (1) indicated a significant prolongation of disease-free survival in favor of the group receiving tamoxifen treatment. Updated results from this study provided by NSABP indicate that the tamoxifen treated patients have a 10-year disease-free survival of 68% compared with 57% for the placebo-treated group (p<0.0001). There is a concomitant improvement in overall survival in favor of tamoxifen, 78% vs. 75% (p=0.037). This beneficial effect of tamoxifen was also noted in other randomized trials (2,3) and has been confirmed by a worldwide meta-analysis of tamoxifen trials performed by the Early Breast Cancer Trialists' Collaborative Group (4).

Although there was an unequivocal prolongation in disease-free survival and survival in favor of tamoxifen, the optimum length of administration of this agent has not been determined. A secondary goal of NSABP B-14 was to compare the effectiveness of 5 years of tamoxifen to 10 years of tamoxifen treatment. Those patients who received 5 years of tamoxifen and were free of disease at the end of this period were re-randomized to either an additional 5 years of tamoxifen or 5 years of placebo.

Results from this re-randomized component of the trial comparing 5 years to 10 years of tamoxifen indicate that there is no additional advantage forthcoming from prolonging the administration of tamoxifen to 10 years. Results from a smaller trial, the Scottish Tamoxifen Trial, also comparing 5 years of tamoxifen to more prolonged treatment, have recently been published in abstract form (5). They are consistent with the conclusion of the NSABP B-14 study.

Complete manuscripts describing details of treatment benefits and side effects from the re-randomized components of both the NSABP and Scottish tamoxifen trials are in preparation. However, because adjuvant tamoxifen is widely used in this country and abroad, rapid disclosure of these data is appropriate. This communication briefly summarizes important features of the NSABP trial and the Scottish trial. Some of the figures may change with the availability of more complete data and longer follow-up.

NSABP B-14

Following either mastectomy or lumpectomy plus breast irradiation for primary breast cancer, women with histologically negative axillary lymph nodes and estrogen receptor-positive tumors were eligible for NSABP Protocol B-14. The study design of B-14 had two components; the first and principal component was initiated in 1981 and consisted of a direct comparison between 5 years of tamoxifen to placebo.

1st Randomization: 1981-88

In the first randomization, 2892 patients were randomized to tamoxifen or placebo. A total of 1439 patients were randomized to tamoxifen (20 mg/d x 5 years) and 1453 patients were randomized to placebo (x 5 years).

This randomization was closed in January 1988 after 2892 patients had been entered. Between January 1988 and October 1988 an additional 1235 patients who fulfilled the same eligibility criteria were registered on this study, all of whom were to receive 5 years of tamoxifen. This patient cohort was added in order to increase the number of tamoxifen-treated patients available for the secondary randomization.

In order to determine whether tamoxifen should be administered for 5 or 10 years following the operation, those patients who received tamoxifen for 5 years (in both the randomized and registered groups) and were disease-free at the end of this 5 year period were offered a second double-blinded randomization to an additional 5 years of tamoxifen or to a placebo.

2nd Randomization: 1987- 1993

In the second randomization, 1166 patients who had received tamoxifen for 5 years were randomized to tamoxifen or placebo. A total of 591 patients were randomized to an additional 5 years of tamoxifen (20 mg/d x 5 years) and 575 patients were randomized to placebo (x 5 years).

B-14 Results:

Results from the initial randomization to tamoxifen versus placebo for 5 years were reported in 1989 (1). They indicate a disease-free survival and overall survival advantage for tamoxifen. Nearly 66% of those eligible, 1172 patients, agreed to the second randomization. Of these, 1166 are eligible for analysis (591 on tamoxifen and 575 on placebo). Starting from the time of the second randomization, the average duration of follow-up is 43 months. Patient pre-treatment characteristics (age, tumor size, type of surgery, ER and PR receptor levels) are well balanced per treatment arm.

Using actuarial estimates at 4 years of follow-up since the re- randomization, the data indicate that 92% of the group that received 5 years of tamoxifen are alive and disease-free compared to 86% of the group scheduled to receive 10 years of tamoxifen. This difference at an interim analysis does not represent statistically significant evidence of a detrimental effect, but it does make it very unlikely that continuation of tamoxifen would produce an actual clinical benefit. As a result of this finding, the trial was stopped by NSABP in concordance with the recommendation of its Data Safety Monitoring Committee. There is no significant difference in overall survival with 96% of the group receiving 5 years of tamoxifen and 94% of the patients receiving 10 years of tamoxifen still alive 4 years after re- randomization. The occurrence of second primary tumors diagnosed after re-randomization included 11 contralateral breast cancers, 5 in the five-year tamoxifen group, 6 in the ten-year group; 8 endometrial cancers, 2 versus 6, and 21 other primaries, 9 versus 12, respectively.

An investigator from St. Luc's Hospital, Montreal submitted falsified data for several NSABP protocols, including NSABP B-14. Among the 1172 patients enrolled in the second randomization, 66 (5.6%) were entered from St. Luc's. When all St. Luc's patients were omitted from the analysis, or when all ineligible patients were omitted, all conclusions remain the same.

B-14 Conclusions:

These results make it very unlikely that continuation of tamoxifen beyond 5 years will result in a disease-free survival or overall survival benefit in women with node-negative, estrogen receptor- positive breast cancers.

THE SCOTTISH TAMOXIFEN TRIAL

Following mastectomy for primary breast cancer, women were randomly allocated to receive adjuvant tamoxifen 20 mg daily for 5 years, or to a control group in which tamoxifen was to be given only on relapse of disease. Like B-14, pre- and postmenopausal women with negative axillary lymph nodes were eligible, but unlike B-14, postmenopausal women with positive axillary nodes were also eligible. In this study, estrogen receptor values were not required. Accrual to the first randomization began in 1978. Subsequently, a second randomization was added in February 1985 that was limited to those women treated on-study with tamoxifen for 5 years who had no recurrence of breast cancer. Thus, only patients who had entered the initial randomization after March 1980 were eligible for the second randomization. Eligible patients were randomized to continue tamoxifen indefinitely or to stop treatment at 5 years.

1st randomization (1978-84)

In the 1st randomization, 1312 patients were randomized to tamoxifen or no treatment for relapse. Of these, 661 were randomized to tamoxifen (20 mg/d x 5 years) and 651 were randomized to no treatment.

2nd randomization (1985-89)

In the 2nd randomization to continue or stop tamoxifen, 173 patients were allocated to continue tamoxifen (20 mg/d) indefinitely and 169 were allocated to stop tamoxifen.

Scottish Trial Results:

The results of the first randomization were reported in 1987 (2) and showed a significant disease-free survival and overall survival benefit for adjuvant tamoxifen. Three hundred and forty-two patients consented to the second randomization, representing 87% of those available for inclusion. Median follow-up for living patients from the date of re-randomization is 6.2 years. Pre-treatment patient characteristics for the re-randomized groups such as age, menopausal status, nodal status, post-mastectomy radiation therapy, and estrogen receptor status were well balanced between the two groups.

At 6.2 years of median follow-up since the re-randomization, the data indicate that 70% of the group that received 5 years of tamoxifen are alive and disease-free compared to 62% of the group that continued treatment beyond 5 years. This represents a non-statistically significant trend in favor of stopping tamoxifen at 5 years. The occurrence of second primary tumors diagnosed after re-randomization included 8 contralateral breast cancers, 3 in the group that stopped tamoxifen at 5 years and 5 in the group continuing tamoxifen indefinitely; 5 endometrial cancers, 0 versus 5, and 18 other primaries, 9 versus 9, respectively.

Scottish Trial Conclusions:

In view of its relatively limited size, this trial is not able to exclude the possibility that continuation of tamoxifen beyond 5 years might be beneficial. However, the data do suggest that, if there is a disease-free survival or overall survival benefit to be derived, it would likely be small. As a result of these findings, no changes have as yet been made regarding the trial policy.

SUMMARY:

Tamoxifen for Breast Cancer Treatment

Adjuvant treatment with tamoxifen results in improved survival for women with early stage breast cancer. The results of NSABP Protocol B- 14 represent an important advance in our understanding of the appropriate duration of tamoxifen treatment. To our knowledge, it is the only available trial with a substantial sample size to directly compare 5 years of tamoxifen treatment to 10 years of tamoxifen treatment. The B-14 data, taken together with the results of the Scottish trial, provide no evidence of benefit for continuing tamoxifen beyond 5 years. The troublesome possibility that continuation of tamoxifen beyond 5 years might actually be detrimental cannot be unequivocally resolved by these data. However, both trials suggest a greater likelihood of relapse in women who take tamoxifen for durations greater than 5 years when compared to women who receive 5 years of therapy.

The NSABP B-14 trial and the Scottish trial highlight the importance to cancer medicine of performing carefully controlled, randomized clinical trials. Obviously, there was no way to anticipate the results of these studies in advance. It is noteworthy that some laboratory investigators hypothesized that prolonged exposure to tamoxifen could induce tumor-dependence (6) or resistance (7).

Although the NSABP and Scottish trials provide evidence that continuation beyond 5 years is not beneficial, the optimal duration of adjuvant tamoxifen treatment still remains to be determined. Indirect evidence from combined analysis of tamoxifen trials suggests that the benefit from 2 to 5 years of treatment is greater than that derived from durations less than 2 years (4). There are currently several trials directly comparing 2 years versus longer durations (3-5 years), but results from these are either not available or not definitive. There are ongoing trials that are also comparing 5 years of treatment versus longer durations. While we eagerly anticipate the results of such trials, all available evidence indicates that 5 years of tamoxifen is a reasonable standard for the adjuvant setting.

Tamoxifen for Breast Cancer Prevention

It is also important to note that NSABP B-14 has served as the basis for the current NCI-sponsored Breast Cancer Prevention Trial (BCPT) in which women who do not have breast cancer but who are at increased risk are randomly allocated to tamoxifen or placebo for 5 years. These new results do not affect the rationale for the BCPT, which is primarily based on the 47% reduction of contralateral breast cancers observed after 5 years of adjuvant tamoxifen therapy. The B- 14 data indicate that 5 years of tamoxifen significantly reduces the incidence of new primary breast cancers in the contralateral breast (52 versus 29 events), and that 10 years of tamoxifen provides no additional benefit over and above this reduction. Because this new information from the B-14 trial does not change the overall risk- benefit profile for 5 years of tamoxifen, NCI supports the continued evaluation of 5 years of tamoxifen as a breast cancer preventive regimen in the context of a randomized clinical trial.

Acknowledgment

The NCI wishes to express its appreciation to the NSABP for providing their data to us and for collaborating on the preparation of this announcement. We also want to thank Dr. Helen J. Stewart and the Scottish Cancer Trials Breast Group for sharing their data with us.

References:

Fisher B, Costantino J, Redmond C, et al: A randomized clinical trial evaluating tamoxifen in the treatment of patients with node- negative breast cancer who have estrogen-receptor-positive tumors. N Engl J Med 1989;320:479-84.
Scottish Cancer Trials Office (MRC): Adjuvant tamoxifen in the management of operable breast cancer: The Scottish trial. Lancet 1987;2:171-74.
Nolvadex Adjuvant Trial Organization: Controlled trial of tamoxifen as a single agent in management of early breast cancer; analysis at six years. Lancet 1985;1:836-39.
Early Breast Cancer Trialists' Collaborative Group: Systemic treatment of early breast cancer by hormonal, cytotoxic, or immune therapy: 133 randomized trials involving 31,000 recurrences and 24,000 deaths among 75,000 women. Lancet 1992;339:1-15, 71- 85.
Stewart H.J. for the Scottish Breast Group: Adjuvant tamoxifen duration in a randomized trial. The Breast 1995; 4(3):256 (abstract 102).
Gottardis MM, Jordan VC: Development of tamoxifen-stimulated growth of MCF-7 tumors in athymic mice after long-term antiestrogen administration. Cancer Res 1988;48:5183-5187.
Sluyser M: Nuclear hormone receptor variants: Their role in malignancy and progression to hormone resistance in cancer. Acta Endocrinol 1991;126 (suppl 1):48-53.

The full text of this alert has been mailed to all libraries that are members of the National Network of Libraries of Medicine.

Last Reviewed: July 31, 2020