DR. I. PERSSON DISCUSSES HORMONE REPLACEMENT AND
Department of Cancer Epidemiology, University of Uppsala,
by Sean Henahan, Access Excellence, America
Even the most complicated scientific studies depend on the
fundamentals of good design and methodology. When the studies
are epidemiological in nature, the picture becomes even more
complex. Recent epidemiologic studies have produced contrasting
results regarding the potential risk of hormone replacement
therapy and risk of breast cancer in menopausal women. How does
one go about sifting the scientific wheat from the statistical
chaff? I asked noted breast cancer researcher and epidemiologist
Ingemaar Persson, M.D., of the department of cancer
epidemiology, University of Uppsala, Sweden, to clarify some of
Could you give us a little information on the reasons estrogen
is suspected of being involved in the development of breast
cancer in the first place?
A:There are numerous pieces of circumstantial
evidence suggesting a possible link between estrogen and breast
cancer. One thing is that the etiology of breast cancer is
associated with reproductive risk factors, such as early
menarche (onset of menses) and late menopause, whereas
protective factors include early age at first birth,
oopherectomy and early menopause. These pieces fit together with
an expectation that ovarian sex steroids would affect breast
cancer risk. Exactly how and when we don't know. We know too
little about the mechanisms of these factors on breast cancer
In addition, steroid hormones are known to regulate the growth
of normal breast epithelium. In some animal studies, estrogens
have been implicated in the development of tumors, but this has
not been seen in all studies.
Q: HRT is a different use of these hormones
than is the case with oral contraceptives. Since HRT has only
been used for some 30 years, is this a long enough period to
even be able to tell if there are any adverse effects?
A:I think so. If you look at the findings in a
large number of studies, you find that there are inconsistent
results, but there are many studies now showing that with many
years of estrogen intake there may be a moderate increase in
risk. This may have to do with the number of years and the
amount of exposure to estrogen.
Q:People reading the newspapers may be confused
to read conflicting reports suggesting an increased or decreased
risk of breast cancer in association with HRT. For example we
recently saw a report from Harvard suggesting a possible
increased risk, and a contrasting report from Seattle not
suggesting increased risk. Can you describe how you reconcile
these kinds of studies?
A:This is a very complicated question. In the
first place, it is wrong to try and make conclusions from only
two studies. In this case, both studies were non-randomized
observational studies. In such studies there are a lot reasons
one might come up with different estimates of risk depending on
the methods used in the studies or shortcomings in those
methods. That is a very real reason we see discrepant results.
It is very, very important when we look at individual studies to
make a judgment on the quality of that study.
Q: Could meta-analysis help clarify the
A:This is very controversial. With
meta-analysis you lump together risk estimates from a large
number of studies. This means you end up combining bad studies
with good studies, qualitatively speaking. That doesn't really
guarantee that you end up with true answers. However, these
studies do offer the advantage of increasing the numbers in the
Q:What about potential confounding factors.
such as differences among estrogens, and the use of combination
therapy involving progestin?
A:In Europe we tend to use estradiol (steroid
containing estrogenic properties) compounds, while in the US you
tend to use the conjugated estrogens (estrogenic compounds
derived from the urine of pregnant mares). These could produce
different effects, but I don't think so. These two common
regimens give about the same effect on endometrial cancer risk,
the same effects on osteoporosis, same effects on coronary heart
disease, and the same effects on relief of menopausal symptoms.
So I don't think there is much difference there.
There could be a difference in another area, the duration and
amount of estrogen intake. This is fundamental. If you are
looking at a relationship between exposure to something and
development of disease, it might be that a specific amount of
exposure is needed for a specific time, to have an effect. That
means if you compare short term users to long-term users, the
results are not comparable.
How can genetic factors be fitted into the risk equation? How
can you control for the portions of the population who may
already be at increased risk?
A:The proportion of women who are assumed to
have the BRCA gene mutations is very small, around five percent.
I don't believe it would affect the results of studies of this
nature. However, another question is whether genetic factors may
play a role in women taking HRT, that is- the phenomenon of
interaction. That is currently hypothesized, , but there are not
very good data, yet. You could examine the presence of genes in
a case-control study.
What we do now is ask women, do you have a mother and/or sister
who developed breast cancer at a young age. If the answer is
yes, there is a higher likelihood that the person may have
genetic predisposition to breast cancer. If you combine first
degree heredity risk for breast cancer with HRT , there are some
hints, although not well established, of an interaction. This is
an interesting area of research.
Q:A number of health benefits have been
attributed to estrogen replacement. These include reduced risk
of heart disease and heart attack and protection against
osteoporosis. How do you calculate the ratio of risks and
benefits in a situation like this?
A:There is some indication of these benefits.
But even here, there is a question of how much these protective
effects could be due to selection. This is a very tricky
problem. It's been shown, in Sweden at least, that women who are
selected by doctors to receive HRT are or who select themselves
to go to the doctor, and who decide to take HRT for a long time
are not 'ordinary average' people. They have higher incomes,
they have higher level of exercise, and smoke less. These
differences could suggest that these women already have a lower
risk for heart disease. Nonetheless, this doesn't mean that
these factors would explain all of the pronounced protective
effects seen with estrogen replacement. in epidemiologic
studies. Taken together, the studies suggest that women taking
HRT have up to a 50% reduction in incidence of heart attack,
compared to those not taking HRT. The same may be true for
Q:In recent years progestins (drugs which mimic
progesterone, a hormone produced by the corpus luteum) have
often been added to estrogen regimens. How does the addition of
progestin change the picture?
A:This is another tricky question. If you are
discussing estrogen and heart disease, this is currently an area
of debate. There may be some adverse effects of progestin on
lipoproteins (cholesterol and related compounds), whereas
estrogens appear to benefit lipids. Some researchers have
wondered if this positive pattern may be reversed with the
addition of progestin. Nobody knows.
The effect of progestin on the risk for breast cancer is another
issue. The Harvard study concluded that there is no difference
in risk relationships between estrogen alone and estrogen plus
progestin. We have seen this in terms of breast cancer in our
Q:What is the message for women who are
contemplating HRT, what should their doctors be telling them
based on the current data?
A:The doctor can tell the patient- there may be
some increased risk of breast cancer in some people, but there
is more reason to think there are potential benefits from HRT.
It is very important to state that short-term use of HRT has no
adverse effects. The risk we are considering would not be seen
until ten to 15 years of estrogen intake. So women who are
having hot flashes or other symptoms of menopause, are in need
of treatment, and estrogen will improve their quality of life
for a period of years.
It is different to consider HRT for prevention of coronary heart
disease or osteoporosis. Both heart attacks and hip fractures
tend to occur in late age. This means you are talking about
long-term treatment, even lifetime treatment. This is where
increased breast cancer risk may manifest. This is where you
have to weigh the potential benefits against potential risks in
- J.L. Stanford et al."...ERT in Relation to Risk of
Breast Cancer in Middle-aged Women", JAMA, 7/13/95.
- I. Persson et al. "HRT and Breast Cancer: A remaining
controversy", JAMA, 7/13/95.
- Brinton et al., "ERT and Breast Cancer Risk", Epidemiologic
Reviews,Vol. 15, No.1.
Related information at other
Frequently Asked Questions about Cancer, University of
Update-Breast Cancer and Hormone Replacement