Breast Paul
C. Stomper,
MD
J. Van
#{149} Bradley
Voorhis,
MD
#{149} Veronica
A. Ravnikar,
MD
Imaging
#{149} Jack
E. Meyer,
MD
Mammographic Changes Associated with Postmenopausal Hormone Replacement Therapy: A Longitudinal Study’
A retrospective study of radiologic findings from 50 postmenopausal women who underwent hormone replacement therapy was performed to determine the frequency and distribution of changes in the screenfilm mammographic appearance of breast tissue between pretreatment and posttreatment mammography. Twelve (24%) of the women developed increased parenchymal density during a median follow-up of 18 months (range, 6-39 months). Mammographic changes included diffuse increase in density in seven (14%) of the women; increased multifocal, asymmetric densities in two (4%); and cyst formation in three (6%) A greater percentage of women who received estrogen-progesterone replacement (26%, 10 of 38) demonstrated changes than those who received estrogen alone (17%, two of 12), but the difference was not statistically significant. The authors conclude that benign increases in mammographic density associated with postmenopausal hormone replacement therapy are not uncommon, and they suggest that the serial changes observed in some women may diminish the sensitivity of mammography for the earlier detection of breast cancer. Index 00.5
terms:
Breast,
#{149} Hormones,
Radiology
1990;
parenchymal
pattern,
00.5
174:487-490
the Department of Radiology, Memorial Institute, School
cine
Biomedical
666 Elm Departments
St.
(B.J.V.V., Brigham ical
VAR.) and and Women’s
School,
vision
(.
Boston.
requested
September
reprint
Science,
Buffalo, NY of Obstetrics
August
RSNA,
1990
to P.C.S.
Rosof Mcdiat Buffalo,
(P.C.S.); and Gynecology
Radiology Hospital,
18; accepted
requests
SUNY
14263 and
Received
URRENT
hormone in
(J.E.M.), Harvard June
2; revision September
30,
the
Mcd1989;
received 22.
Address
re-
medical opinion replacement
menopausal
women,
and
MATERIALS
favors therapy
This
such
literature
that
compared
pretreat-
and
AND
study
women
therapy is being widely prescribed. Proved benefits of estrogen replacement therapy include the prevention of vasomotor symptoms, genital atrophy, and osteoporosis (1-3). Possible adverse effects include an increased risk of endometrial cancer unless estrogen is combined with progestenone therapy (4). Recent case-control studies show no increased risk of breast cancer in women taking hormone replacement therapy (5,6). The mole of progesterone in possibly decreasing breast cancer risks is controversial (7,8). Peck and Lowman (9) first discussed the observation that postmenopausal estrogen replacement therapy could produce benign changes in the breast panenchyma demonstrable by mammography and that, following cessation of hormone therapy, regression of these changes could be seen radiologically. They suggested that in certain patients, estrogen therapy should be discontinued to enhance the nadiologic as well as clinical detection of early breast cancer. Bland et al (10), in a matchedcohort study of 405 postmenopausal patients, did not show a significant difference in xemomammographic parenchymal patterns between patients treated with hormones and untreated patients. We know of no studies in the
From well Park and
C
seen
breast
of 50 consecutive
at a specialized screening
menopause
clinic
of the
Brigham
and
Women’s
Hospital,
who
the
following
criteria:
met
woman both
was by
that and
shown
symptoms
menopausal
and
mange;
screen-film
by
within
hormone
ists
and (d)
in
pen-
to the
therapy
start
and
of
at ap-
1-year intervals thereafter; replacement therapy was maintained
throughout
and
was
prior
replacement
tiated
hormone were
values
(b) high-resolution
1 year
proximately (c) hormone
Each
determination
mammography
formed
Boston, (a)
to be postmenopausa!
serum follicle-stimulating luteinizing hormone
by
the
the
clinical
course
patients
had
of
no
mi-
special-
the
study;
prior
breast
cancer, significant alteration in body weight during the period of the study, concurrent systemic illness, breast-related symptoms,
on history
on had on
not
undergone
was
52 years
surgery.
tients
The
of thyroid
disorder,
breast
median
irradiation
age
of
(range,
the
pa-
40-69
years).
Seventeen (34%) of the women were aged 40-49 years; 33 (66%) were aged 50 years or greater.
Hormone replacement therapy ed of estrogen alone in 12 patients combination estrogen-progesterone placement
in
38 patients.
consistand re-
Estrogen
thera-
consisted of 0.05 mg of transdenmal tradiol every 3 days in 27 women, 0.625 mg of conjugated equine estrogen each
es-
py
day
in eight
cnonized
women,
and
estradiol
Progesterone
each
therapy
13-25
of the
the
cycle
in
eight
ment
(Senographe
CGR
USA,
Paris;
2.5
acetate
on each
women, acetate on
and each
mg
of
day
1 .0 mg day of
obtained
screen-film
by
us-
equip-
500T,
General
Ontho-M
film
screens, Eastman-Kodak, With the exception
10 mg
on days
27 women,
women. were
high-resolution
of miof
acetate
cycle in three All mammograms
ing
mg
in 15 women.
consisted
medmoxypmogestenone in
1.0
day
of medroxypnogestenone
of the cycle of nonethindrone
ment and posttreatment mammograms to determine the changes seen over time in a series of women undergoing hormone therapy. This study was undertaken to determine the frequency and distnibution of changes in the screen-film mammographic appearance of breast
METHODS
consisted
Electric
and
Rochester, of five pretreatment
Min-R NY).
undergoing
mammognaphic studies performed at outside centers, all mammography was performed within the clinic. The mammograms were interpreted by
therapy.
two mammogmaphems mammographic
tissue
group
before
and
after
treatment
in
of carefully monitored women estrogen or combination estrogen-progesterone replacement
a
(P.C.S., panenchymal
J.E.M.). pattern
The on
487
a.
b.
Figure 18
1.
(a)
months
Pretreatment
mammograms
of estrogen-progesterone
of a 52-year-old replacement
woman
therapy
show
show diffuse
a. (a) Pretreatment mammograms of a 69-year-old woman breast panenchyma. (b) Mammograms obtained after 13 estrogen-progesterone replacement therapy show diffuse density. (c) Mammograms obtained after 27 months of
treatment
show
pretreatment oblique
progressive
bilateral projections
was retrospectively lows: dense (greater sue), fatty (greaten on heterogeneous tions). The anatomic relations of dense
have
been
field
(1 1). The
described
mammogmaphica!!y tively
categorized
serial
diffuse
increase
change);
488
pamenchyma.
(b)
Mammograms
obtained
after
density.
means
.
categorized as fo!than 90% dense tisthan 90% fatty tissue), (all
other combinaand pathologic mammographic
contissue
by Page
and
changes
observed
were
then
as follows:
on cyst development of ultrasound [US]).
Radiology
in density.
craniocaudal of each woman
Winc.
retrospecno
change;
diffuse increase in density (at least 10% change); increased multifoca!, asymmetric densities that did not exhibit mass effect or architectural distortion (at least 10% by
breast
in
b.
Figure 2. show fatty months of increase in
the and
heterogeneous increase
(confirmed The follow-
up
period
tween
was
the
start
defined
acquisition of the most ment mammognam. Ninety-five percent vals and statistical
as the
of hormone
interval
recent
were
and
posttneat-
confidence
Fisher exact tests analysis (12).
be-
therapy
interused
for
RESULTS The mammographic panenchymal pattern on the pretreatment mammogram was categorized as dense in nine (18%) of the women, fatty in six (12%), and heterogeneous in the re-
February
1990
Figure
3.
months
(a)
Pretreatment
mammograms
of estrogen-progesterone
replacement
a. Figure
of
a 47-year-old
therapy
woman
show
show
increased
b. 4
(a) Pretreatment
mammogram
fatty
breast
multifocal,
panenchyma.
asymmetric
woman
ed no change. tnibution
Volume
174
#{149} Number
2
(arrows).
of
heterogeneous
breast
on the mammograms interval, 13%-38%). 38 (76%) demonstratThe these
obtained
after
29
d.
shows
chymal density (95% confidence The remaining
35 (70%). During a median period of 18 months 6-39 months), 12 (24%) of 50 developed increased pamen-
Mammograms
c. of a 42-year-old
12 months of estrogen replacement alone shows formation of circumscribed masses. (c) US scan (arrow). The more superficial lesion demonstrated increased sound transmission on other images. of treatment shows resolution of the posterior cyst and enlargement of the anterior cyst (arrow),
maining follow-up (range, women
(b)
densities
frequency changes
and are
dis-
out-
lined in the Table and illustrated in Figures 1-4. Thirty-two (64%) of the women underwent one follow-up mammographic study only. Ten of the 12 women who exhibited serial changes underwent mammography at approximately 1 year and 2 years after the start of therapy. Six (60%) demonstrated change at 1 year; four (40%) first demonstrated change 2 years after therapy. Two of 12 (17%) women receiving estrogen therapy alone demonstrated serial change over a median followup period of 15 months (mange, 6-25 months). Ten (26%) of 38 women me-
parenchyma.
(b) Mammogmam
demonstrates the (d) Mammogram which was again
presence obtained confirmed
obtained
after
of simple cysts after 26 months with US.
ceiving combination estrogen-progestemone therapy demonstrated semial change over a median follow-up period of 18 months (range, 6-39 months). A Fisher exact test showed no significant difference between these percentages (P .70). No breast cancers or mammographic abnormalities warranting biopsy were found in this group of senial screening mammograms.
DISCUSSION It is generally accepted that the mammognaphic density of breast tissue decreases or remains stable after menopause, and that cyst formation is a prernenopausal
(13). breast
Although parenchymal
phenomenon
cohort
studies patterns
Radiology
of show #{149} 489
denser breast tissue patterns in youngem and premenopausal women cornpared with older and postmenopausal women (14,15), little published data show the changes toward less dense breast panenchymal patterns that occur with aging. A series meported by Wolfe (16) showed that such changes generally occur between the ages of 35 and 50 years. In light of this historical experience and the exclusion of known factors that could alter breast density (ie, breast surgery, significant weight loss, mastitis), we believe that the observed serial changes in our postmenopausal study population can be attributed to the effects of hormone replacement therapy. Our observation that 24% of treated women exhibited conversion to mammographically denser breast tissue within a relatively short median follow-up period of 18 months shows that this is not an uncommon occurnence. The observations that four patients first demonstrated changes on the second annual posttreatment mammogram and that two patients (Figs 2, 4) demonstrated increasing density on two consecutive posttreatment mammograms suggest that the frequency and degree of mammographic panenchymal changes is time-rebated and will increase with the duration of hormone replacement therapy. Although the frequency of mammographic changes in women treated with estrogen meplacement alone (17%) was less than in women treated with combined estrogen-progesterone replacement (26%), there was no statistically significant difference. The numbers in this study are too small for the findings to be conclusive. We believe that knowledge of these panenchymal patterns of change (development of diffuse density, asymmetric densities, or cysts) and awareness of the use of hormone replacement therapy by women undergoing screening mammography should prevent mammographers from diagnosing benign changes related to hormone therapy as develop-
490
#{149} Radiology
ment of asymmetric densities on inflammatony change suggestive of malignancy. Careful, shorter-interval, clinical and mammognaphic followup may be appropriate in some cases. However, demonstration of other signs of malignancy (ie, solid masses, architectural distortion, microcalcifications, on skin changes) warrants the usual suspicion and further evaluation. The effect of these honmone-related changes on the sensitivity of mammography for the earlier detection of breast cancer has not been determined. Data from the Breast Cancer Detection Demonstration Project showed a sensitivity of approximately 93% for mammography at the time of screening in women aged 50 years or more, compared with 85% in wornen aged 40-49 years (17). This is usuably attributed to the denser breast parenchyma in younger women and associated potential for obscuration of a soft-tissue, noncalcified, mabignant mass that does not cause appreciable distortion of the surrounding normal breast tissue. For the small percentage of women who exhibit significant increase in breast density during postmenopausal hormone meplacement therapy (Figs 2, 4), the sensitivity of mammognaphic detection of early breast cancer may be diminished. This would occur in older women, who have a greater risk of breast cancer than younger, premenopausal women, who usually have dense parenchymal patterns (18). We believe that for this apparently small subset of women undergoing honmone replacement therapy, it may be prudent for the mammographer to suggest that the referring physician and patient reevaluate the risks and benefits of the patient’s hormone meplacement regimen and consider withdrawing therapy to enhance the nadiologic detection of early breast cancer. U
References 1.
ica!
assistance.
James Frigoletto,
We
Talcott, MD,
We thank Catherine Elliott for their invaluable
are
grateful
MD,
Judy
and
for
Garber,
Kenneth
the
input
MD, Ryan.
No-
of
Frederick MD.
analysis
prod 2.
considerations climactenic:
of prevalent
Med
1984;
Semmens
JP,
in a dinJ Re-
treatments.
29:71-79.
Wagner
gen therapy menopause. 18.
G.
Effects
of estno-
on vaginal physiology Obstet Gynecol 1985;
Hammond
4.
tus of estrogen therapy for the menopause. Fertil Stenil 1982; 37:5-25. Gambrell RD Jr. The menopause: benefits and risks of estrogen-progestogen replacement therapy. Fertil Stenil 1982; 37:457-
5.
CB,
473. Wingo
PA,
Maxson
Lazde
WS.
during 66:15-
3.
PM,
Lee
Current
NC,
sta-
Rubin
G,
Ory HW. The risk of breast cancen in postmenopausal women who have used estrogen replacement therapy. JAMA 1987;
257:209-215.
6.
Kaufman DW, Miller DR. Lonenbeng L, et al. Noncontraceptive estrogen use and the risk of breast cancer. JAMA 1984;
7.
Gambrel! creased
252:63-69.
RD Jr. Maien R, Sancens incidence of breast cancer
menopausal 8.
BI. Dein post-
estrogen-progestogen
Obstet
Gynecol
World
Health
1983;
users.
62:435.
Organization.
Collabora-
tive study of neoplasia ceptives, breast cancer, and medroxyprogesterone
and stenoid contracervical cancer, acetate. Lancet
1984; 2:1207-1208.
9.
Peck
DR.
Lowman
RM.
postmenopausa!
Estrogen
breast:
considerations.
JAMA
and
the
mammognaphic
1978;
240:1733-
1735.
10.
Bland KI, Buchaan JB, Weisberg BF, et a!. The effects of exogenous estrogen replacement therapy of the breast: breast cancer risk
and
mammographic
patterns.
11. 12.
13. 14.
15.
Cancer
panenchyma! 1980;
45:3027-3033.
Page DL, Winfield AC. The dense mogram. AJR 1986; 147:487-489.
mam-
Fisher
ne-
RA.
Statistical
methods
for
search workers. Edinburgh: Edinburgh University Press, 1934. Hoeffker W, Lanyi M. Mammography. Philadelphia: Saunders, 1977; 78-88. Grove JS, Goodman MJ, Gilbert Fl, Mi MP. Factors associated with mammographic pattern. Br J Radio! 1985; 58:21-25. Leinster
SJ,
Walsh
PV,
Sumidare AM. mammographic Clin
Radio!
1988;
Wolfe JN. Breast and their changes 1976;
121:545-552.
17.
Baker
LH.
port. Horm
1982;
JW,
cancer
Program:
tality
in the
85-1837.
of Health,
AJ,
United
Bethesda,
year
dem-
summary
ne-
32:194-225. Young
cancer
Institutes
a!-
patterns Radiology
detection
five
Mine
SEER tional
parenchyma! with age.
project:
CA
GH,
39:252-256.
Breast
onstnation
Whitehouse
Factors associated with panenchymal patterns.
16.
18.
Acknowledgments: yak, RI, and Carol
Upton GV. Therapeutic the management of the
JL,
Pollack
ES.
and
mom-
incidence States,
of Health Md:
1973-81.
Na-
Publication National
no.
Institutes
1984.
February
1990