Reference
Cuzick J, Sestak I, Forbes JF, et al. Anastrozole for prevention of breast cancer in high-risk postmenopausal women (IBIS-II): an international, double-blind, randomised placebo-controlled trial. Lancet. 2013. [Epub ahead of print]
Design
Between 2003 and 2012, postmenopausal women were enrolled in a double-blind, randomized placebo-controlled trial comparing anastrozole versus placebo for 5 years.
Participants
Women at high risk for breast cancer, aged 40–70 years, were recruited from 18 countries. The women were randomly assigned to receive either the study medication (n=1,920) or placebo (n=1,944). Those at high risk for breast cancer were defined as “women aged 45–60 years who had a relative risk of breast cancer that was at least 2 times higher than in the general population, those aged 60–70 years who had a risk that was at least 1.5 times higher, and those aged 40–44 years who had a risk that was 4 times higher.”
Study Medication and Dosage
1 mg oral anastrozole (Arimidex) or matching placebo every day for 5 years
Outcome Measures
The primary endpoint was confirmed breast cancer.
Key Findings
After a median follow-up of 5.0 years, 40 women in the anastrozole group (2%) and 85 in the placebo group (4%) had developed breast cancer (HR: 0.47, 95% CI: 0.32–0.68, P<0.0001).
The predicted cumulative incidence of all breast cancers after 7 years was 5.6% in the placebo group and 2.8% in the anastrozole group. Eighteen deaths were reported in the anastrozole group and 17 in the placebo group, and no specific causes of death were more common in 1 group than the other (P=0.836).
Clinical Implications
The decision to review this study requires explanation, as anastrozole is hardly "natural medicine." This paper is of importance for several reasons; first it may foretell a time when the standard of care suggests that women at high risk for breast cancer be prescribed anastrozole and our patients will be asking us for advice. Second, there are natural substances that have aromatase inhibitory action and data from this paper may encourage the use of these materials by certain patients.
One of the study’s authors, Jack Cuzick, PhD, declared at last winter’s San Antonio Breast Cancer Symposium, "Our results provide substantial support for the use of anastrozole as the treatment of first choice for prevention of breast cancer in high-risk postmenopausal women."
Some will always see the cup half full while others will see it half empty. We should surely question this optimism and point out that these numbers translate into a number needed to treat of 36; that is, it’s estimated that treating 36 women for 5 years will prevent 1 case of breast cancer in 7 years. This seems like a whole lot of drugs to have a mild impact.
We must find and consider the bottom line. Eighteen deaths were reported in the anastrozole group and 17 in the placebo group. At this point 5 years into the experiment there is no difference in overall survival (OS), and the data do not yet support the assumption that there will be a fewer deaths from breast cancer in the anastrozole group with longer follow-up.
There is an assumption that reduction in cancers that are diagnosed, or improved progression-free survival (PFS) as seen in this trial, will correlate with improved OS numbers; however this is not actually the case. In fact in most cancers, and specifically in breast cancer, using PFS as a surrogate marker for OS is not supported in the literature.
So what does this trial actually show us? Cutting risk of cancer by half (53%) sounds striking, but it is somewhat misleading because the change is out of context. We lack information on prevalence. Risk for those taking placebo was 5.6%; the drug cut that risk in half to 2.8%. Put another way, the risk of breast cancer for a patient who declines to take anastrozole if it were offered, would, after 7 years, increase by 1 in 36, or about 2.8%. That sounds different than saying her risk will double. It is also important to keep the OS data in mind; in those 5 years her risk of dying remains the same whether taking anastrozole or placebo.
Something else of note in these data was that the incidence of some other cancers also decreased in the women taking anastrozole. Not counting breast cancer, there were a total of 40 cancers in the women taking the drug (2%) compared to 70 in those taking placebo: a 42% reduction [RR=0.58 (0.39–0.85)].
For example, there were 14 skin cancers in the treatment group compared to 27 in the placebo group, a risk ratio of 0.53 (0.28–0.99). There were 3 colorectal cancers in the treatment group compared to 11 in the placebo group, a risk ratio of 0.28 (0.08–0.99). (See Table 1)
Although the anastrozole group had fewer cases of breast cancer, it had more breast cancer deaths.
The study authors see this as an “exciting the possibility—that this drug may have an effect in reducing the risk of other cancer.” Our patients on the other hand are not likely to be excited about taking anastrozole, since lowering estrogen levels can come with the side effects associated with estrogen deprivation.
Again, it is important to keep in mind the difference between PFS and OS. Although the prevalence of cancer was decreased in the anastrozole group, the death rates between the groups were the same. Interestingly, although the anastrozole group had fewer deaths from other cancers, it actually had more breast cancer deaths then the placebo group. So here we have the juxtaposition between PFS and OS laid out for us very clearly: Although the anastrozole group had fewer cases of breast cancer, it had more breast cancer deaths. Although treatment with anastrozole decreases the number of woman who will get a breast cancer recurrence in a given time period, those that do recur are more deadly. (See Table 2)
There are other data from this trial that are noteworthy. The drug was surprisingly well tolerated, with only about 4% difference in compliance in drug vs placebo groups. About 64% of patients treated with the drug complained of musculoskeletal pain, which is similar to other previous estimates, but interestingly 58% of women taking placebo reported musculoskeletal pain. Similar ratios were reported for those with moderate pain (22% with anastrozole, 19% with placebo), while severe pain was reported by 8% of women in the anastrozole group, compared with 6% on placebo.
Cuzick explained, "There’s a general perception that the aromatase inhibitors are very toxic and cause a lot of aches and pains. But 90% of the aches and pains in this study had nothing to do with the drug. They’re just a reflection of the fact that in the postmenopausal years women do get aches and pains."
This has significant clinical implication as we have thought that about 70% of women taking aromatase inhibitors will have significant pain and it may be that this belief has encouraged blame to fall on the drug when it isn’t always the culprit.
This information might still influence our patient recommendations, as there are a number of natural aromatase inhibitors that we might encourage patients to consume as foods or as supplements. Flaxseed meal and white button mushrooms both can be eaten as foods and both inhibit aromatase activity. Melatonin and grape seed extract may be the most common nutritional supplements with aromatase-inhibiting properties.
Thus while the data in this study may not push us to prescribe these drugs to prevent cancer, they do provide a basis for us to encourage consumption of natural aromatase inhibitors in high-risk women or women already diagnosed with cancers that were reduced in this trial.
For more research involving integrative oncology, click here.
Table 1
Cancer Occurrence in Anastrozole vs Placebo Groups
| Type of Cancer | Anastrozole (n=1,920) | Placebo (n=1,944) | Risk ratio (95% CI) |
| Skin cancer | 14 (1%) | 27 (1%) | 0.53 (0.28–0.99) |
| -Non-melanoma | 10 (1%) | 20 (1%) | 0.51 (0.24–1.08) |
| -Melanoma | 4 (<1%) | 7 (<1%) | 0.58 (0.17–1.97) |
| Gastrointestinal | 4 (<1%) | 12 (1%) | 0.34 (0.11–1.04) |
| Colorectal | 3 (<1%) | 11 (1%) | 0.28 (0.08–0.99) |
| Endometrial cancer | 3 (<1%) | 5 (<1%) | 0.61 (0.15–2.54) |
| Leukaemia, lymphoma, or myeloma | 4 (<1%) | 7 (<1%) | 0.58 (0.17–1.97) |
| Thyroid cancer | 0 2 (<1%) .. | ||
| Urinary tract | 2 (<1%) | 5 (<1%) | 0.41 (0.08–2.08) |
| Nervous system | 3 (<1%) 0 .. | ||
| Lung cancer | 4 (<1%) | 4 (<1%) | 1.01 (0.25–4.04) |
| Ovarian cancer | 4 (<1%) | 7 (<1%) | 0.58 (0.17–1.97) |
| Vaginal cancer | 1 (<1%) 0 .. | ||
| Carcinomatosis | 1 (<1%) | 1 (<1%) | 1.01 (0.06–16.18) |
| Total* | 40 (2%) | 70 (4%) | 0.58 (0.39–0.85) |
Table 2
Causes of Death in Each Group
| Cause of Death | Anastrozole group (n=1920) | Placebo group (n=1944) |
| Breast cancer | 2 (<1%) | 0 |
| Other cancer | 7 (<1%) | 10 (1%) |
| Cerebrovascular accident or stroke | 2 (<1%) | 2 (<1%) |
| Cardiac arrest | 3 (<1%) | 1 (<1%) |
| Other | 4 (<1%) | 4 (<1%) |
| Total | 18 (1%) | 17 (1%) |
