Dixon Questions Value of Chemoprevention With Tamoxifen

J. Michael Dixon, MD, OBE

The reduction of breast cancer risk with 5 years of tamoxifen does not outweigh the lack of a mortality benefit, according to J. Michael Dixon, MD, OBE, who discussed his position at the 32nd Annual Miami Breast Cancer Conference on Sunday.

In long-term, follow-up data (16-22 years) from the IBIS-1 trial presented at SABCS in December, 5 years of tamoxifen was shown to reduce the risk of breast cancer by 29% in otherwise healthy women at high risk of the disease; however, tamoxifen continued to have no impact on breast cancer mortality. At 20 years’ follow-up, there were 31 breast cancer—related deaths in the tamoxifen arm versus 26 with placebo.

In an interview with Oncology Nursing News, Dixon, who is professor of Surgery, consultant surgeon, and clinical director of the Breakthrough Research Unit at the University of Edinburgh in Scotland, said he considers these results “sobering.”

“If you can prevent breast cancers but can’t prevent people from dying of breast cancer, it’s a point for discussion, because patients are taking 5 years of a drug that has significant side effects,” said Dixon. “If you don’t prevent people from dying, then how valuable is that? I’m not sure it’s that valuable.”

IBIS-I Basics

The ongoing IBIS-1 study, known formally as the International Breast Cancer Intervention Study-1, recruited 7154 women at breast care and genetics clinics in eight countries between 1992 and 2001 and randomized them evenly to receive 20 mg of daily tamoxifen for 5 years (n = 3579) or matching placebo (n = 3575).

The median age of participants across both cohorts was 50.8 years; slightly more than 50% were postmenopausal, and the mean body mass index was 27. Women with an increased risk of breast cancer, largely due to a family history of the disease, were eligible for the study regardless of whether they were using hormone-replacement therapy (HRT) before the trial, during it, or had not used HRT, or whether they had undergone a hysterectomy. These characteristics were also balanced across the two arms.

“Between 40% and 50% of patients in the two arms took HRT during the trial and nowadays you wouldn’t get many women on HRT. Nevertheless, 50% of patients in the two groups were never-users of HRT and about 35% of women on each arm had had a hysterectomy,” said Dixon.

At 10 years of follow-up, substantial differences between the two groups were observed confirming previously reported analyses. One-hundred sixty-three breast cancers were reported in the tamoxifen arm (4.6%), versus 226 (6.3%) with placebo (HR = 0.72; 95% CI, 0.59-0.88). For invasive ER-positive disease specifically, 100 breast cancers were reported in the treatment arm versus 145 in controls (HR = 0.68; 95% CI, 0.53-0.88). For all breast cancer, the number needed to treat (NNT) for 5 years to prevent one breast cancer was 59 .

At 20 years’ follow-up, there was a 29% reduction in breast cancer risk with tamoxifen (HR = 0.71; 95% CI, 0.60-0.83; P <.0001). A total of 601 breast cancers were reported. In the tamoxifen cohort, breast cancer was reported in 251 women versus 350 in the placebo arm. For invasive ER-positive breast cancer, risk was reduced by 35%; 160 cases were reported in the treatment arm versus 238 with placebo (HR = 0.66; 95% CI, 0.54-0.81; P = .0001). Dixon did note, however, that “In the second 10 years, there is a suggestion—it’s not significant, though&mdash;that you get more invasive ER-negative cancers.”

Although the overall risk reduction remained comparable in the second 10 years, the NNT dropped to 22. Also, the difference between breast cancer incidence rates between the two arms grew: 7.8% in the tamoxifen arm versus 12.3% in the placebo arm. For invasive ER-positive breast cancer, NNT was 29 and incidence rates were 4.9% versus 8.3%.

The effects were larger for women who did not receive HRT during the trial, at 38% versus 12%, respectfully. “If you took HRT during the trial, you didn’t get much of a reduction in the incidence of breast cancer. You got a much bigger decrease in the incidence of breast cancer in the women who had never taken HRT,” said Dixon.

“So, after 20 years, it’s quite clear that you can prevent breast cancer, and if you look at the numbers needed to treat, you only need to treat 22 women to stop 1 breast cancer from developing, 29 to stop 1 ER-positive invasive cancer, and the effects are largest for the women not taking HRT,” said Dixon.

With non-breast cancers, there were 36 more incidents in the tamoxifen arm (n = 351; 9.8%) compared with the placebo arm (n = 315; 8.8%). The gap was due to occurrences of non-melanoma skin cancers, at 116 versus 84, and endometrial cancers at 29 versus 20. With the combined incidence of all other cancers, there were slightly fewer in the tamoxifen arm.

“There was an excess of other cancers (HR of 1.13—not quite significant) in the tamoxifen arm. And one of the reasons is that there’s an excess of endometrial cancer&mdash;don’t forget a lot of these women had had hysterectomies. But the excess of endometrial cancer is all limited to the first 5 years. You had 15 endometrial cancers in the tamoxifen arm and 4 in the placebo arm in the first 5 years,” said Dixon.

Among notable adverse events, the incidence of pulmonary embolism (PE; 30 vs 22) and deep venous thrombosis without PE (40 vs 23) was higher in the tamoxifen versus the placebo arm. More patients receiving placebo had myocardial infarction (n = 17) than those in the tamoxifen group (n = 13).

No Mortality Benefit

Despite the prevention benefit observed at 20 years, tamoxifen chemoprevention continued to have no benefit on breast cancer mortality. There were 5 more breast cancer mortalities in the tamoxifen arm (n = 31) than in the placebo arm (n = 26).

“The big concern is that there is no reduction in breast cancer mortality, and there’s a slight increase of deaths seen after 10 years of tamoxifen….It’s quite concerning that in the second 10 years, in the placebo arm there were 9 deaths from breast cancer and in the tamoxifen arm, there were 18,” said Dixon.

There were also 5 endometrial cancer mortalities with tamoxifen versus zero for placebo, and deaths from all cancers were 83 versus 78 in the two arms, respectively.

Although chemoprevention with tamoxifen is not common practice in the United Kingdom, Dixon shared his opinion on how he would advise patients.

“I’m not sure from what I’ve seen that I would be encouraging new patients who come along to take tamoxifen. I think you’ve either got the very-high risk women who need prophylactic mastectomy or the moderate risk women who need intensive screening, and I think this may be a big blow for chemoprevention.”

There are those in the breast cancer community who disagree with Dixon’s view, including the lead author of IBIS-1, Jack Cuzick, PhD, who presented the results at SABCS in December. Cuzick, the John Snow professor of Epidemiology at the Wolfson Institute of Preventive Medicine at Queen Mary University of London, considers it a “major issue” that chemoprevention with tamoxifen is not used more frequently worldwide.

“It’s a major issue. I think cardiologists have been very effective in terms of identifying things like high blood pressure and high cholesterol as diseases that need treatment. We have to find a way to make it very clear that women at high risk of breast cancer should be offered treatment much more often,” Cuzick said in a press briefing at SABCS.

Still, Dixon is strong in his conviction. “Ultimately we have to prevent deaths from breast cancer, as well as preventing breast cancers.”

A Possible Explanation?

In his presentation today, Dixon will also discuss a potential explanation for the lack of mortality benefit with tamoxifen in the IBIS-1 trial, which is rooted in the research described in his abstract from SABCS, “In-depth genomic analysis of ER+ breast cancers during development of endocrine resistance.”²

Dixon et al’s genomic assessment included 17 postmenopausal women with ER+ breast cancer who received neoadjuvant letrozole. Four patients responded well and 13 progressed on treatment or had an initial response and then developed acquired resistance.

Dixon said the researchers found that “over time, the responding cancers seem to lose mutations…as though we’re ‘knocking out,’ the mutant cells.” However, in cancers that developed acquired resistance, “there are certain mutations present when [the tumor] is growing actively that weren’t present at all before that. It’s a different cancer at the end than what it was at the start.”

This same premise may explain what’s occurring in the IBIS-1 tamoxifen population, according to Dixon.

“It may be that tamoxifen is actually breeding out cancers that are biologically more aggressive because they’ve got these new mutations,” said Dixon, “You knock off mutations in sensitive cancers, but you get new mutations in nonresponding cancers, and it may start to explain why you, obviously, prevent some breast cancers but others are developing cancer cells that are more biologically aggressive, and so it is not resulting in an overall improvement in mortality.”

References

• Cuzick J, Sestak I, Cawthorn S, et al. Extended long-term follow up of the IBIS-I breast cancer prevention trial. Presented at: San Antonio Breast Cancer Symposium; December 9-13, 2014; San Antonio, TX. Abstract S3-07.
• Dixon JM, Turnbull AK, Fan C, et al. In-depth genomic analysis of ER+ breast cancers during development of endocrine resistance Presented at: San Antonio Breast Cancer Symposium; December 9-13, 2014; San Antonio, TX. Abstract S1-05.