The authors of the 2003 Million Women Study were wrong to conclude their investigation proved that hormone replacement therapy caused cancer, according to a series of articles published this week in the Journal of Family Planning and Reproductive Health.
Endocrinologist and paper co-author Professor Henry Burger explains that there may be a modest increase in cancer risk for some hormone therapy users but it’s important menopausal women get clear and accurate information about the risks and benefits of this therapy:
During menopause, the ovaries effectively stop making estrogen. Over the two to three year lead-up to a woman’s final spontaneous menstrual period, the estrogen levels in her blood have dropped by around 90%.
Around one in five women don’t experience any symptoms. But the same proportion suffers from severe symptoms such as hot flushes, night sweats (which destroy the quality of her sleep) and vaginal dryness (which makes intercourse uncomfortable).
Hormone replacement therapy aims to give back some estrogen to prevent or relieve these symptoms. It’s 90 to 95% effective, which is far greater than other treatments.
What did the 2003 Million Women Study claim about the link between the use of hormone replacement therapy and cancer?
This study of one million women who presented for routine mammograms in the United Kingdom asked participants to complete a questionnaire about hormone therapy and the risk of breast cancer. The researchers looked at the rate at which breast cancer was diagnosed in these women and their history of using hormone replacement therapy.
The authors claimed the study definitively proved hormone therapy caused breast cancer. And this is the focus of the papers we released this week, led by Professor Samuel Shapiro (I was a co-author). We asked whether the Million Women Study authors came to a valid conclusion.
Our analysis found it did not meet a number of requirements for a valid epidemiological study and therefore the authors were not justified in claiming this was the final proof that hormone therapy caused breast cancer.
It’s important to be clear that we’re not disputing the possibility of a link between hormone therapy and breast cancer. What we’re disputing is the use of the Million Women Study data to prove that link. The Million Women Study is what’s called an observational study, which cannot prove causality – it can only prove association.
What impact did the 2003 study have on the use of hormone therapy?
It’s difficult to say because not long before that, the Women’s Health Initiative (WHI) published a randomised control trial of hormone therapy, which resulted in a profound drop in the use of this therapy – around 80% in the United States.
There’s no doubt the Million Women Study added to the fear and anxiety generated by the American WHI study.
What is known about the actual links between hormone therapy and cancer?
It’s important to put any risk on the context of the overall analysis of benefit and risk.
If you’re looking at the benefits, hormone therapy can relieve or cure the symptoms of menopause in the 20% of women who have severe symptoms. It’s also suitable for women who have moderate symptoms and would rather have them treated. So 30 to 40% of menopausal women could benefit from the therapy.
The second benefit it a reduction in the risk of osteoporotic fracture. Around the time of menopause, with the fall in estrogen, there’s a loss of bone tissue, which increases the risk of fracture. Hormone therapy has been shown to prevent or substantially reduce that loss of bone and therefore reduce the risk of fracture.
A third demonstrated benefit is that oral therapy reduces the risk of colorectal cancer. It actually reduced the number of colorectal cancers in the WHI study by almost the same number as the increase in breast cancers.
The WHI study also showed there was a reduction in the incidence of diabetes. And there’s also growing evidence to show it may reduce her risk of dementia and heart disease.
What are the risks of hormone therapy?
There is a small increase in the risk of Venous thromboembolism, if the hormone is given orally. This isn’t present if it’s given through the skin via an estrogen patch.
There’s also a small increase in the risk of breast cancer. That risk varies depending on the duration of the therapy, the type of therapy and her other risk factors.
As an endocrinologist, how do you tackle this issue of balancing risk and benefits with your patients?
When I see patients with severe symptoms of menopause, I describe the benefits and the risks. I tell her there is a lot of evidence to suggest hormone therapy increases the risk of breast cancer, then we’ll look at her individual risk. If she’s a very slim lady, then her risks are probably a bit higher. Likewise, if she has a family history of breast cancer, her baseline risk will be slightly higher.
I’ll tell her it’s unlikely that hormone therapy will increase her risk of breast cancer if she uses it for less than five years.
I also tell her that in order to put that risk in perspective we can look at other things that increase her risk of breast cancer – being overweight or obese confers a higher risk of breast cancer than hormone therapy. As does an early first period (before 12 years), a late menopause, and childlessness. Hormone therapy falls into the lower end of a series of risk factors, which can increase a woman’s risk of breast cancer.
I tell her she has to weigh all of this up for herself in terms of her perception of how bad her symptoms are versus how much that risk is. And most women will say this isn’t a risk that worries them and they’re happy to take it.
What advice would you give to regulatory authorities about revising their clinical guidelines for prescribing hormone therapy?
Regulatory authorities in Australia and around the world used the WHI and the Million Women Study to formulate their policies and advise clinicians and women about the use and duration of hormone therapy. The conclusions of these studies aren’t valid so the regulatory authorities must acknowledge this and adjust their advice accordingly.