Editorial: Refining Risk Assessment in Women With Benign Breast Disease: An Ongoing Dilemma
JNCI J Natl Cancer Inst (
Refining Risk Assessment in Women With Benign Breast Disease: An Ongoing Dilemma
Stuart J. Schnitt 0 1
Monica Morrow 0 1
Nadine M. Tung 0 1
0 The Author 2017. Published by Oxford University Press. All rights reserved. For Permissions , please
1 Affiliations of authors: Department of Pathology (SJS) and Division of Hematology-Oncology (NMT), Beth Israel Deaconess Medical Center and Harvard Medical School , Boston , MA; Department of Surgery, Memorial Sloan Kettering Cancer Center , New York, NY, MM , USA
-
More than three decades ago, Dupont and Page published their
seminal study relating breast cancer risk to the histologic
findings in benign breast biopsies (
1
). The key observations of that
study were that women whose biopsy showed proliferative
lesions without atypia had about a twofold increase in the risk of
subsequent breast cancer and those with atypical hyperplasia
had about a fivefold increase in breast cancer risk when
compared with women with nonproliferative lesions. As
summarized in Table 1, subsequent studies from other groups have
yielded strikingly similar findings, despite differences in the
study design, patient populations, and pathologists involved in
the histologic classification of the benign breast biopsies (
1–5
).
As a result of these studies, which in aggregate have included
more than 17 000 women, atypical hyperplasia is considered a
high-risk lesion and is included in risk assessment models such
as the Gail/Breast Cancer Risk Assessment Tool (BCRAT) (
6,7
)
and the International Breast Cancer Intervention Study (IBIS)
model (8); proliferative lesions without atypia are also included
in the IBIS model. Further, the finding of proliferative lesions in
a benign breast biopsy, especially atypical hyperplasia, can
result in more intensive screening and pharmacologic risk
reduction with selective estrogen receptor modulators or aromatase
inhibitors (
9
).
In this issue of the Journal, Visscher et al. report on the risk
of breast cancer among a small subset of women in the Mayo
Clinic Benign Breast Disease (BBD) cohort who had multiple
metachronous benign breast biopsies (
10
). Among 1414 women
who had more than one biopsy (10.5% of their study
population), there was a change in BBD category between the first and
subsequent biopsy in 43.9% of cases. Further, the authors
observed that a change from a lower-risk to a higher-risk BBD
category between the first and subsequent biopsy was associated
with an increase in breast cancer risk whereas a change from a
higher-risk to a lower-risk BBD category was associated with a
reduction in risk. What are the implications of these findings
for clinical management, and what insights do they provide
regarding the biology of breast tumor progression?
In our view, the clinical management issues are
straightforward for women whose second biopsy shows a higher-risk
category of BBD than the initial biopsy, particularly for those
women whose subsequent biopsy shows atypical hyperplasia.
According to the data of Visscher et al., these women are at
least the same level of increased risk as any other women with
atypical hyperplasia (if not higher) and they should be managed
accordingly (
10
).
The most interesting group is the subset of women who had
atypical hyperplasia in the first biopsy but had nonproliferative
lesions or proliferative lesions without atypia in the subsequent
biopsy. The data of Visscher et al. suggest that these women
have a 43% reduction in breast cancer risk when compared with
women who still have atypical hyperplasia in their second
biopsy (
10
). However, this scenario appears to be extremely
uncommon; 42 such women were identified in this study,
representing only 3% of women with multiple biopsies and 0.3%
of the entire Mayo Clinic BBD Cohort (
10
). Further, before
concluding that these patients are no longer at high risk of breast
cancer, it should be noted that this reduction in risk was not
statistically significant. In addition, these women continue to
have over a threefold increase in their risk of breast cancer
(hazard ratio ¼ 3.36, 95% confidence interval ¼ 1.34 to 8.45) when
compared with women who had nonproliferative lesions in
both their first and subsequent biopsies. Finally, given the
persistently elevated breast cancer risk in this group, it is possible
that atypical hyperplasia is in fact present in the breast tissue,
but in areas other than that subjected to the subsequent biopsy.
We view this as plausible because more than half of the
subsequent biopsies in this population were performed for clinically
detectable lumps or mammographic masses (Supplementary
*CI ¼ confidence interval; RR ¼ relative risk.
Nonproliferative
disease
RR (95% CI)
Proliferative disease
without atypia
RR (95% CI)
1 (ref)
1 (ref)
Table 3, available online) (
10
); the finding of atypical hyperplasia
in such specimens is much less common than in biopsies
performed because of mammographic (...truncated)