CHICAGOUntil findings from two recent clinical studies were
released, there was no clear indication about the effectiveness of
tamoxifen (Nolvadex) or prophylactic mastectomy for reducing the risk
of invasive breast cancer in high-risk women.
However, initial results from the NSABP P-1 trial 2 years ago showed
that a 5-year course of tamoxifen could reduce the incidence of
invasive breast cancer by 49% and ductal carcinoma in situ (DCIS) by
50% among women at increased risk for breast cancer.
A 1999 retrospective study of moderate- and high-risk women who were
followed for 14 years after undergoing prophylactic mastectomy at the
Mayo Clinic indicated that the surgical procedure reduced breast
cancer incidence and death by 90% (Hartmann LC et al: N Engl J Med
Nevertheless, no single preventive strategy is appropriate for every
woman at increased risk for breast cancer. Both tamoxifen and
prophylactic mastectomy have their uses in highly selected
patients, Monica Morrow, MD, said at the Second Annual Lynn
Sage Breast Cancer Symposium.
These preventive options also have associated risks, such as an
increased incidence of endometrial cancer, thrombotic events,
postsurgical complications, and reoperations.
Clinicians therefore need to integrate breast cancer prevention into
their practices by selecting a prophylactic measure in keeping with
the magnitude of a womans risk, the potential side effects of
the treatment option, and the timing of the intervention, said Dr.
Morrow, professor of surgery and director of the Lynn Sage
Comprehensive Breast Center, Northwestern University Medical School.
Dr. Morrow pointed out that making preventive care decisions on the
basis of breast cancer risk was difficult in the past because of
disagreement about the role of individual risk factors as well as
their dependence on age and their interactions with one another.
The Gail Model
The Gail model, which was used in the NSABP P-1 study of tamoxifen in
the chemoprevention of breast cancer, has resolved some of these
concerns because it considers a womans age, age at menarche,
age at first birth, number of first- degree relatives with breast
cancer, history of previous breast biopsies, and history of atypical
The Gail model, in fact, was one of the most useful pieces of
information to come out of the NSABP tamoxifen prevention trial, Dr.
Morrow said. During the 5 years of the study, the Gail model
predicted that 159 invasive breast cancers would occur, and 155 did
The model underestimated risk for women in the lowest risk group by
about 30%, and it overestimated risk in the highest risk group by
20%. However, none of the over- or underestimates reached traditional
statistical significance according to confidence intervals.
Something to keep in mind is that the Gail model is an
estimate, not an absolute number, and at either end of the spectrum
of risk, there is room for variability, Dr. Morrow said.
The Gail model also was accurate in predicting the risk of invasive
breast cancer on the basis of some individual risk factors, such as
age at menarche. The model was less successful in determining risk
for young women who had two or more breast biopsies.
For women over the age of 50, the model accurately predicted breast
cancer risk; the rate of expected-to-observed cases (E/O) was 1.07
when atypical hyperplasia was absent. For women under the age of 50,
it overestimated the development of breast cancer whether atypical
hyperplasia was present (E/O 2.00) or absent (E/O 1.76).
This finding is not surprising, Dr. Morrow said, because breast
biopsy is no longer restricted to clinically apparent mass lesions;
some clinicians count cyst aspiration as a biopsy or insert biopsy
needles into bumpy areas to reassure both the patient and themselves
that no cancer is present.
In addition, improvements in mammography are revealing more
suspicious lesions that are subjected to image-guided biopsy.
Clinicians consequently should be wary when applying the Gail model
to younger women whose major component of breast cancer risk is a
history of multiple breast biopsies.
The number of breast biopsies may very well reflect the
patients anxiety or the physicians anxiety about the
level of risk as much as any biological process, she said.
The Gail model should not be applied to other patient populations,
such as women with a personal history of breast cancer, DCIS, or
lobular carcinoma in situ, or a strong family history suggestive of a
genetic predisposition to breast cancer.
From the NSABP, we know that tamoxifen reduces risk due to a
family history, histological precursors, and other risk factors, but
we do not have data to say whether or not it reduces risk due to
genetic mutation, Dr. Morrow said. There is no definitive
information to say that tamoxifen works, and no information to say it
Finally, she said, the Gail model should be used with caution when
assessing the risk of breast cancer in nonwhite women. Only 99 black
women were included in the placebo arm of the NSABP P-1 trial, and
the number of women from other racial groups was even smaller.
The decision to use tamoxifen in breast cancer prevention must
balance the chance that it may reduce the risk of invasive disease
against the potential for causing life-threatening side effects, such
as endometrial cancer and thrombotic vascular events.
Risk of Endometrial Cancer
The NSABP P-1 trial found a 2.5 overall relative risk of endometrial
cancer in postmenopausal women taking tamox-ifen. A case-controlled
study refined the assessment of endometrial cancer risk from
tamoxifen according to hormone exposure and body mass index (BMI)
(Bernstein ML et al: J Natl Cancer Inst 91:1654-1662, 1999).
This study of 324 women who were treated with tamoxifen for breast
cancer and 671 matched controls showed an overall 1.5 rate of
increased risk of endometrial cancer among women who used tamoxifen.
The risk of endometrial cancer was greater among women who had taken
estrogen previously (3.53) than among those who had not (1.14), and
it was highest among women who had taken estrogen previously and had
a high BMI (3.95).
It is the obese woman with a prior history of estrogen use who
has a substantial elevation in her risk of endometrial cancer,
Dr. Morrow said.
For the woman with low BMI and no prior estrogen use, she pointed
out, there was no statistically significant elevation of risk in this
study. So we can say that tamoxifen definitely increases the
risk of endometrial cancer, and patients need to be informed about
that. But this risk needs to be modulated by other known risk
factors, and in talking to women, we need to try to individualize our
estimates of risk, Dr. Morrow said.
The other important complication of tamoxifen use is thrombotic
vascular events, which appears to be age related; the only
statistically significant increase in risk for stroke, pulmonary
embolism, and deep-vein thrombosis occurs in women over the age of
50, she observed.
Findings from risk-benefit studies, are beginning to suggest when
tamoxifen may be used to reduce the risk of breast cancer. The
findings dont mean that every premenopausal woman with a
risk level for breast cancer greater than 1.5 needs to be
treated, she said. It means that the likelihood of
significant complications is extremely low.
The choice for a postmenopausal woman seems to revolve around the
issue of whether she has a functioning uterus. As a woman gets
older, she said, she needs a higher level of breast
cancer risk to offset the risk of deep-vein thrombosis and
endometrial cancer. So knowing whether or not a woman has her uterus
is a clinically important piece of information in postmenopausal women.
According to the Hartmann study, prophylactic mastectomy may be
appropriate for women at high risk of breast cancer because of a
strong family history of the disease, which suggests a genetic
predisposition. However, the procedure is associated with surgical
Also, the first breast reconstruction after mastectomy often is not
the last operation. The incidence of unplanned reoperations within 5
years is about 30% for implants and 10% for tissue flap reconstruction.
Dr. Morrow noted that not all women at increased risk for
breast cancer require treatment, nor will they all benefit from
treatment because of other comor-bid conditions. Not all women who
will benefit from treatment want treatment. We must regard risk as a
personal decision. Undertaking risk reduction is something a women
does when she finds her level of risk to be unacceptable, not
necessarily when we find her risk to be unacceptable.