21 Years Later: Adjuvant Hormonal Therapy for Elderly Breast Cancer Patients

Abstract & Commentary

Synopsis: In an adjuvant breast cancer trial conducted more than 20 years ago, older patients were randomized to treatment with tamoxifen and prednisone or no adjuvant therapy. Now, at 21 years of follow-up, it is clear that disease-free and overall survival remain better for those who received treatment. Current practice, however, uses a more protracted course of tamoxifen, and it is possible that long-term results for older patients under treatment may reflect a larger number of adverse outcomes.

Source: Crivellari D, et al. J Clin Oncol. 2003;21: 4517-4523.

Breast cancer occurs most commonly in older women, and it is currently observed more frequently in early stages. Yet, the matter of adjuvant therapy remains unsettled for this age group. In the current report, the data from the International Breast Study Group (Trial IV) after 21 years of follow-up is presented. This multi-institutional trial, which was conducted from 1978 to 1981, randomized lymph node-positive patients, aged 66-80, who were randomly assigned to treatment with tamoxifen plus low-dose prednisone (p+T) or no adjuvant therapy. Of the 349 enrolled, there were 329 evaluable patients. Demographic characteristics were equally divided in the 2 treatment groups.

After a median follow-up of 21 years, a single year of p+T significantly prolonged disease-free survival (DFS; P = .003) and overall survival ([P = .05]; 15-year DFS, 10% ± 3% vs 19% ± 3%; hazard ratio, 0.71; 95% CI, 0.58-0.86). The DFS advantage was seen for patients in the following subpopulations; 1-3 positive nodes (P = .04), 4 or more nodes (P = .04), ER-positive patients (P = .02), patients with unknown ER status (P = .009), tumor size > 2 cm (P = .03), age 66-69 years (P = .01), and age 70-80 years (P = .04) but not separately for ER-negative and smaller tumor size subgroups.

When comparing competing causes of failure (breast cancer recurrence and deaths before breast cancer recurrence), p+T was far superior in controlling breast cancer recurrence (P = .0003), but the improvement was seen mainly in soft-tissue sites. Conversely, patients in the p+T group were more likely to die before a breast cancer recurrence (P = .03).

Comment by William B. Ershler, MD

Such long-term success of adjuvant therapy in elderly breast cancer patients had not been previously reported. Data from the same study were published at a median follow-up of 8 years,1 and it is encouraging to find the initial positive results were sustained. In 1978 when this trial was initiated, the optimal duration of therapy was unclear and a concurrent ECOG trial (examined 2 years of tamoxifen vs placebo in a similar population of patients) also resulted in favorable early findings.2,3 Most consensus guidelines currently recommend a 5-year course of tamoxifen for node-positive ER+ patients.4,5 The long-term results for older patients treated in this manner may demonstrate that deaths from causes other than breast cancer are increased, presumably due to increased thromboembolic disease associated with tamoxifen treatment, for which the elderly appear particularly susceptible (ref30). However, such may not be the case with the equally effective, but less toxic aromatase inhibitors.6

Thus, tamoxifen has been shown to enhance disease free and overall survival, even when administered for 1 year only (with prednisone) in elderly node positive patients. It is probable that more sustained treatment (eg, 5 years) will also have a positive effect on long-term outcomes, but these will have to be balanced with a likely increase in encountered adverse events.  

Dr. Ershler is Fairfax Hospital Cancer Center, Fairfax, VA; Director, Institute for Advanced Studies in Aging, Washington, DC.


1. Castiglione M, et al. J Clin Oncol. 1990;8:519-526.

2. Cummings FJ, et al. Ann Intern Med. 1985;103: 324-329.

3. Cummings FJ, et al. J Clin Oncol. 1993;11:3-4.

4. NIH Consensus Development Conference. J Natl Cancer Inst. 2001;93:979-989.

5. Goldhirsch A, et al. J Clin Oncol. 2001;19: 3817-3827.

6. ATAC Trialists Group. Lancet. 2002;359:2131-2139.