Cookies on this website

We use cookies to ensure that we give you the best experience on our website. If you click 'Accept all cookies' we'll assume that you are happy to receive all cookies and you won't see this message again. If you click 'Reject all non-essential cookies' only necessary cookies providing core functionality such as security, network management, and accessibility will be enabled. Click 'Find out more' for information on how to change your cookie settings.

Accept all cookies Reject all non-essential cookies Find out more

9.4 The third cycle (1995)

In the third cycle of the EBCTCG, the number of eligible trials was expanded to include all those that began before 1990. Information was obtained on approximately 37 000 women who had been randomized in 55 trials of adjuvant tamoxifen versus no tamoxifen, comprising about 87% of all the women randomized into eligible trials (EBCTCG 1998b). Compared with the second cycle, the amount of evidence on events occurring more than five years after randomization was substantially increased and, in terms of numbers of deaths, the amount of evidence from trials of about five years of tamoxifen was doubled. Nearly 8000 of the randomized women had a low, or zero, level of oestrogen-receptor protein measured in their primary tumour; these are ‘ER-negative’ tumours.

Among these women the overall effects of tamoxifen were small, with the annual mortality rate reduced by only 6% (4%), and there was no suggestion of any trend towards greater benefit with longer treatment. In contrast, among the remaining women, of whom there were 18 000 with ER-positive tumours and nearly 12,000 women with untested tumours, the results were striking: for trials of one year, two years, and five years of adjuvant tamoxifen, mortality was reduced by 12% (3%), 17% (3%), and 26% (4%), respectively, during ten years of follow-up. Not only were all these reductions highly significant individually, but there was some evidence of a greater effect with longer treatment (p = 0.003). The proportionate mortality reductions were similar for women with node-positive and node-negative disease, but in absolute terms the reductions were greater in node-positive women, in whom survival is generally poorer.

In trials with the longest course of treatment, of about five years of tamoxifen, the absolute improvements in ten-year survival were 11% (2%) for node-positive (61% versus 50%; p < 0.00001) and 6% (1%) for node-negative (79% versus 73%; p < 0.00001). These benefits were largely irrespective of age, menopausal status, daily tamoxifen dose, and of whether or not the women in both arms of the trial had received chemotherapy. In terms of other outcomes, about five years of tamoxifen approximately quadrupled the incidence of endometrial cancer (p = 0.0008), and halved the incidence of cancer in the contralateral breast p < 0.00001): in absolute terms, however, the decrease in the incidence of contralateral breast cancer was about twice as large as the increase in the incidence of endometrial cancer. Tamoxifen had no apparent effect on the incidence of colorectal cancer or, after exclusion of deaths from breast or endometrial cancer, on any of the other main categories of cause of death. About one extra death per 5000 woman-years of tamoxifen was attributed to pulmonary embolus but, based on the evidence available at that time, the excess was not statistically significant.

Information was available in the third cycle of the EBCTCG on about 18,000 women randomized in 47 trials of prolonged polychemotherapy versus no chemotherapy, about 6000 women randomized in 11 trials of longer versus shorter polychemotherapy, and about 6000 women in 11 trials of anthracyclinecontaining regimens versus cyclophosphamide, methotrexate, and fluorouracil (EBCTCG 1998a). In trials of polychemotherapy versus no chemotherapy there were highly significant reductions in mortality both for women aged under 50, for whom there was a 27% (5%) reduction (p < 0.00001), and for women aged 50–69 at randomization, who showed an 11% (3%) reduction (p = 0.0001), while few women aged 70 or over had been studied. After taking both age and time since randomization into account, the proportionate reductions in mortality were similar in women with node-negative and node-positive disease. These proportionate reductions suggest that, for women aged under 50 at randomization, the effect of polychemotherapy would be to increase a typical ten-year survival of 71% for those with node-negative disease to 78%, giving an absolute benefit of 7%, while for women with node-positive disease it would be to increase a typical ten-year survival of 42% to 53%, an absolute benefit of 11%. For women aged 50–69 at randomization, the corresponding increases would be from 67% to 69% for those with node-negative disease, an absolute benefit of only 2%, and 46% to 49% for those with node-positive disease, giving an absolute benefit of only 3%. At a given age, the benefits of polychemotherapy appeared largely independent of menopausal status at presentation, ER status of the primary tumour, and of whether or not adjuvant tamoxifen had been given. In addition, the directly randomized comparisons of polychemotherapy did not indicate any survival advantage with the use of more than about six months of polychemotherapy. In contrast, the directly randomised comparisons provided some evidence that, compared with cyclophosphamide, methotrexate, and fluorouracil alone, the anthracycline-containing regimens reduced mortality slightly, to a five-year survival of 69% versus 72% (p = 0.02). In terms of other endpoints, polychemotherapy reduced the incidence of cancer in the contralateral breast by about one-fifth and had no apparent adverse effect on deaths fromcauses other than breast cancer. Although the EBCTCG data show that polychemotherapy can improve long-term survival, it does have considerable shortterm side effects, including leukopenia, nausea and vomiting, thromboembolic events, thrombocytopenia, anaemia, infection, mucositis, diarrhoea, and neurological toxicity (Pritchard et al. 1997, Fisher et al. 1990). The incidence of such events has not been reviewed by the EBCTCG, but clearly needs to be taken into account in treatment decisions.

Trials of the effect of ovarian ablation were also considered in the third cycle of the EBCTCG (EBCTCG 1996). Many of these trials began before 1980, and so 15 years of follow-up were available for analysis. Among the 1354 women aged 50 or over when randomized, there was no significant improvement of ovarian ablation on survival, as in the results obtained in the second cycle. However, among the 2102 women aged 50 or under when randomized, 15-year survival was highly significantly improved among those allocated ovarian ablation—52.4% versus 46.1%, so 6.3% (2.3%) fewer deaths per 100 women treated—and the benefit was significant both for node-positive and for node-negative disease. In the trials of ablation plus cytotoxic chemotherapy versus the same chemotherapy alone, the benefit of ablation appeared smaller than in the trials in the absence of chemotherapy, and although there was no significant heterogeneity between these two subgroups, the benefit of ovarian ablation on survival was not significant when the trials in which both arms had received chemotherapy were considered on their own. It was concluded that further randomized evidence was needed on the additional effect of ovarian ablation in the presence of other adjuvant treatments and also on the relevance of hormone receptor measurements, which were only available for four of the 12 trials.

Many of the trials comparing radiotherapy plus surgery with the same type of surgery alone started even earlier than the trials of ovarian ablation. Therefore, in the third cycle of the EBCTCG, it was possible to examine the impact of radiotherapy at both ten and at twenty years (EBCTCG 2000). Data were available for a total of 40 trials, involving 20,000 women, half with node-positive disease. In these trials the radiotherapy fields included not only chest wall or residual breast, but also the axillary, supraclavicular, and internal mammary chain lymph nodes. Breast cancer mortality was reduced in the women who had received radiotherapy (p = 0.0001) but mortality from certain other causes, in particular cardiovascular disease, was increased, and overall 20-year survival was 37.1% with radiotherapy versus 35.9% without (p = 0.06). Nodal status, age and decade of follow-up strongly affected the ratio of breast cancer mortality to other mortality, and hence affected the ratio of absolute benefit to absolute hazard. It was estimated that, without the long-term hazard, radiotherapy would have produced an absolute increase in 20-year survival of about 2–4%, except for women at particularly low risk of local recurrence. The average hazard seen in these trials would, however, reduce this 20-year survival benefit in young women and reverse it in older women.

Until recently, it had been thought that an increased risk of cardiovascular disease occurred only following substantial doses of radiation, and improvements in radiotherapy techniques for breast cancer in recent years have tended to reduce radiation doses to the heart. However, there is now mounting evidence that doses that have traditionally been regarded by clinical oncologists as unimportant in terms of cardiovascular risk may, in fact, carry an appreciable long-term risk. The main evidence for this comes from the follow-up studies of the survivors of the atomic bombings of Hiroshima and Nagasaki, where significant dose-response relationships for heart disease and for stroke have been reported following whole body uniform doses (Preston et al. 2003). These data also suggest that if there is a threshold dose for non-cancer disease mortality then it cannot be greater than about 0.5 Gy and this suggests that, even with modern radiotherapy techniques, some cardiovascular risk may well remain. Further research is now being carried out, using data from cancer registries (Darby et al. 2003) together with detailed dosimetry data, to characterize more precisely the cardiovascular risk from radiotherapy for breast cancer.