Upload
julia-a-knight
View
216
Download
0
Embed Size (px)
Citation preview
EPIDEMIOLOGY
Ovarian cysts and breast cancer: results from the Women’sContraceptive and Reproductive Experiences Study
Julia A. Knight Æ Maia Lesosky Æ Kristina M. Blackmore Æ Lynda F. Voigt ÆVictoria L. Holt Æ Leslie Bernstein Æ Polly A. Marchbanks Æ Ronald T. Burkman ÆJanet R. Daling Æ Alice S. Whittemore
Received: 25 April 2007 / Accepted: 30 May 2007 / Published online: 7 July 2007
� Springer Science+Business Media B.V. 2007
Abstract A diagnosis of ovarian cysts is likely an
indicator of hormonal milieu and thus may be related to
breast cancer risk. Recent studies have reported an inverse
relationship between prior ovarian cyst diagnosis and
breast cancer risk. We evaluated this relationship in the
Women’s Contraceptive and Reproductive Experiences
(CARE) Study, a population-based case–control study
conducted in Atlanta, Detroit, Philadelphia, Los Angeles,
and Seattle. Cases had first primary invasive breast cancer
diagnosed between 1994 and 1998 at ages 35–64 years.
African American women were over-sampled. Controls
were identified through random digit dialling and were
frequency matched to cases on centre, race, and five-year
age group. A total of 4575 cases and 4682 controls were
interviewed. We used unconditional logistic regression
adjusted for age and study centre within racial groups to
estimate the odds ratio (OR) and 95% confidence interval
(CI) for the relationship between prior ovarian cysts and
breast cancer. Ovarian cyst diagnosis was associated with
a significantly reduced risk among Caucasians (OR =
0.85, 95% CI 0.76–0.96) and among African Americans
(OR = 0.68, 95% CI 0.57–0.81). The association in
Caucasians was not significant within subgroups defined
by menopausal status, hormone use, or gynecological
surgery while the OR estimates in African Americans
were consistently lower and frequently significant. These
data are consistent with the previously reported inverse
association between ovarian cysts and breast cancer, but
the evidence for a relationship was stronger in African
Americans than Caucasians. Additional studies are
required to determine the specific cyst type(s) responsible
for the observed relationship.
Keywords Breast cancer � Ovarian cysts � Case–control
study � Race � Hormone receptor status
Introduction
The hypothesis that a history of ovarian cysts is associated
with breast cancer risk is plausible as these cysts are likely
an indicator of hormonal milieu. Initial reports were largely
null [1–3], but two recent studies with larger samples have
reported a reduced risk of breast cancer significantly
associated with a diagnosis of ovarian cysts [4, 5].
J. A. Knight (&) � M. Lesosky � K. M. Blackmore
Prosserman Centre for Health Research, Samuel Lunenfeld
Research Institute, Mount Sinai Hospital, 60 Murray Street, Box
18, Toronto, ON, Canada M5T 3L9
e-mail: [email protected]
L. F. Voigt � V. L. Holt � J. R. Daling
Division of Public Health Sciences, Fred Hutchinson Cancer
Research Centre, Seattle, WA, USA
L. F. Voigt � V. L. Holt
Department of Epidemiology, University of Washington, Seattle,
WA, USA
L. Bernstein
Department of Preventive Medicine, Keck School of Medicine,
University of Southern California, Los Angeles, CA, USA
P. A. Marchbanks
Division of Reproductive Health, Centers for Disease Control
and Prevention, Atlanta, GA, USA
R. T. Burkman
Department of Obstetrics and Gynecology, Bay State Medical
Center, Springfield, MA, USA
A. S. Whittemore
Department of Health Research and Policy, Stanford University
School of Medicine, Stanford, CA, USA
123
Breast Cancer Res Treat (2008) 109:157–164
DOI 10.1007/s10549-007-9634-4
A diagnosis of ovarian cysts can arise from a number of
cyst types. Functional cysts, which may or may not be
symptomatic, are the most common type and they frequently
resolve without intervention. Less common are endometri-
omas occurring as a manifestation of endometriosis. Benign
neoplastic tumours such as cystic teratomas and serous
cystadenomas are sometimes grouped with other ovarian
cyst types. Polycystic ovary syndrome (PCOS), with multi-
ple small cysts associated with high levels of androgens, has
been evaluated as a potential risk factor for breast cancer
with mixed results with findings of both increased and
decreased risk as well as null results [6–9].
The etiology of ovarian cysts and benign ovarian tumours
is not well understood and evidence suggests that the risk
factors for the various entities that make up this group of
conditions differ [10]. In addition, some ‘‘risk factors’’ such
as menstrual irregularity, impaired fertility, and oral con-
traceptive use may occur as a result of the condition rather
than as a predisposing factor. Oral contraceptives have been
used to prevent functional cysts, although modern, low-dose
oral contraceptives do not appear to affect the risk of
developing ovarian cysts [11, 12]. However, functional and/
or other types of cysts may be a marker of a particular state of
the hypothalamic-pituitary-gonadal axis that is associated
with reduced breast cancer risk.
Here, we report results from the Women’s Contraceptive
and Reproductive Experiences (CARE) Study, a large,
multi-site case–control study of breast cancer conducted in
the US. This study collected information on ovarian cysts,
ovarian surgery, and other breast cancer risk factors prior to
breast cancer from 4,575 cases and 4,682 controls, pro-
viding an additional opportunity to evaluate a potential
inverse relationship between ovarian cysts and breast
cancer risk.
Material and methods
Study design and population
The Women’s CARE Study, which is funded by the
National Institute of Child Health and Human Develop-
ment, is a population-based case–control study conducted
in Atlanta, Detroit, Philadelphia, Los Angeles, and Seattle.
The study design has been described in detail elsewhere
[13]. Protocols were approved by the institutional review
boards at each site and all the women provided a signed
informed consent form. The study was restricted to Cau-
casian and African American women able to speak English
who were born in the US.
Cases were randomly sampled from women with inci-
dent, histologically confirmed, first primary invasive breast
cancer diagnosed between July 1, 1994 and April 30, 1998
at ages 35–64 years who resided within the geographic
area covered by each participating centre. At all centres
except Philadelphia, women were identified through local
cancer registries that participated in the Surveillance,
Epidemiology, and End Results (SEER) Program. In
Philadelphia cases were identified through area medical
facilities by study personnel. Younger cases and African
American cases were over-sampled to approximate a
uniform distribution across age groups and groups of
Caucasian and African American women. Of the 5,982
eligible cases identified, 4,575 (76.5%) were interviewed.
The controls were women without invasive or in situ
breast cancer who were living in the same geographic loca-
tions as the cases. Random-digit dialling of residential
households was used to identify eligible controls. Through-
out the study, women were randomly sampled from the
eligible controls in order to frequency match the controls to
the cases on study centre, race, and five-year age group.
Approximately 82% of residential households contacted
were successfully screened. Of the 5,956 eligible women
selected as controls, 4,682 (78.6%) were interviewed.
In-person interviews were conducted within 18 months
(most were within six months) of the reference date, which
was defined as date of diagnosis for cases and date of first
household contact for controls. Information was collected
on reproductive and medical history, family history of
cancer, and demographic and lifestyle variables. Respon-
dents were asked whether they had ever had a diagnosis of
ovarian cysts before the reference date and a complete
ascertainment of ovarian and other gynecological surgeries
was obtained.
Statistical analysis
Analyses were based on the 4,555 cases and 4767 controls
with information on a diagnosis of ovarian cysts. Reference
age was defined as age at diagnosis for cases and age at first
household contact for controls. We estimated odds ratios
(ORs) and 95% confidence intervals (CIs) using uncondi-
tional logistic regression. The main variable of interest was
ever having been diagnosed with ovarian cysts prior to the
reference date. Information on type of cysts was not col-
lected. All analyses were performed separately for Cau-
casians and African Americans as the interaction between
cysts and race was statistically significant (P = 0.03). All
models included reference age as a continuous variable and
also study centre (Atlanta, Detroit, Philadelphia, Los
Angeles, and Seattle). We individually tested the following
variables as potential confounders: Hispanic ancestry (yes/
no), education (three categories), oral contraceptive use
(ever/never), breastfeeding (at least two weeks/never or
less than two weeks), hormone therapy (HT) use (never and
ever any estrogen plus progestin, estrogen only, progestin
158 Breast Cancer Res Treat (2008) 109:157–164
123
only), first degree family history of breast cancer (yes/no),
menopausal status (premenopausal, postmenopausal,
unknown), menopausal status with age at menopause
(premenopausal, unknown, and age at menopause in four
categories), body mass index (BMI, kg/m2) five years prior
to the reference date (four categories), BMI at age 18 years
(three categories), age at menarche (three categories),
income (five categories), parity (five categories including
nulliparous), age at first birth (five categories including
nulliparous), alcohol use (never/former/current), smoking
(never/former/current), hysterectomy (ever/never), number
of ovaries removed (none, one, both). None of these vari-
ables altered the OR for ovarian cysts by 10% or more in
either race except for number of ovaries removed in
Caucasians, which increased the OR by 10%. We then
estimated ovarian cyst ORs within each of the following
subgroups: menopausal status (premenopausal and post-
menopausal), HT (never, ever any estrogen plus progestin,
estrogen only, progestin only), hysterectomy (never/ever),
ovaries removed (none, one, two). The associations
between breast cancer and age at cyst diagnosis (no cysts,
age <30 years, age ‡30 years) and time since cyst diag-
nosis (no cysts, <20 years, ‡20 years) were also estimated.
Finally we estimated the ORs for ovarian cysts within
subsets of cases based on estrogen receptor (ER) and
progesterone receptor (PR) status using all controls. All
analyses were carried out using SAS version 9 and with a
two-tailed P-value <0.05 to define statistical significance.
Results
Table 1 describes cases and controls separately among the
5,954 Caucasians (2,938 cases and 3,016 controls) and
among the 3,277 African Americans (1,617 cases and
1,660 controls). Cases and controls were similar for a
number of factors including age, income, and education. As
expected, controls tended to have a higher parity and they
were also more likely to have breastfed. Cases were less
likely to have had a hysterectomy or ovaries removed.
Among Caucasians, ER status was available for 90% of
cases and PR status was available for 87% of cases. The
proportion was somewhat lower for African American
cases in whom ER was available for 81% and PR was
available for 76%.
Among all Caucasians and among all African Americans
ovarian cysts were statistically significantly associated with a
reduced risk of breast cancer, although the effect was greater
among African Americans (Table 2). Among Caucasians,
the OR estimate was less than 1.0 in all subgroups (according
to menopausal status, HT use, hysterectomy status, and
oophorectomy status), but was not more than marginally
statistically significant in any subgroup. In contrast, African
Americans had consistently lower OR estimates and these
were significantly below 1.0 in a number of subgroups. We
found no statistically significant interactions between cysts
and menopausal status, HT use, hysterectomy, or removal of
ovaries in either racial group. There was also no evidence of a
difference between cysts diagnosed before age 30 and those
diagnosed at an older age. Among Caucasians, a statistically
significant inverse association with cysts was only observed
for cysts diagnosed 20 years or more before the reference
date, but time since the reference date did not have much
impact in African Americans. From Table 3 it appears that
the difference between the races is confined to ER or PR
positive breast cancers as there was little difference in the
relationship of cysts to either ER or PR negative tumours,
although none of the differences reach statistical significance
(P = 0.10 and 0.15 for ER positive and PR positive,
respectively and P = 0.87 and 0.88 for ER negative and PR
negative, respectively).
Discussion
In this analysis we found some evidence, using the
Women’s CARE Study, to support the hypothesis that a
history of ovarian cysts is associated with a reduced risk of
breast cancer. Consistent with recent reports from two
other large studies [4, 5], we observed significant inverse
associations between a reported diagnosis of ovarian cysts
and breast cancer. However, in the present study the evi-
dence for an association was stronger in African Americans
than in Caucasians. The relationship between prior ovarian
cysts and breast cancer in African Americans has not
previously been evaluated in the published literature.
Initial reports relating ovarian cysts to breast cancer risk
were largely null. A case–control study of 354 cases and
747 controls identified from women undergoing breast
screening in the US had an OR of 1.1 (95% CI 0.7–1.6) for
a report of cystic ovary [1]. Another US case–control study
of 2,173 cases and 1990 controls under age 55 observed an
OR of 1.16 (95% CI 0.8–1.7) for a history of ovarian cyst
treated with surgery [3]. The risk was significantly elevated
in those who had surgery for the cysts, but had not had
oophorectomy. In a cohort study of women who had
undergone oophorectomy and/or hysterectomy in Sweden
295 cases of breast cancer occurred and the observed
standardized incidence ratio was 0.8 (95% CI 0.5–1.5) for
having ovarian cyst as an indication for surgery [2].
Recently, data from Caucasians from three study sites in
the US, Canada, and Australia were pooled yielding 3,049
cases, 2,344 population controls, and 1,934 sister controls
[4]. An OR of 0.71 (95% CI 0.61–0.81) was observed for a
reported diagnosis of ovarian cysts and the result was
consistent across study sites and was also consistent
Breast Cancer Res Treat (2008) 109:157–164 159
123
Table 1 Characteristics of Caucasian and African American cases and controls with information on ovarian cysts from the CARE study
Caucasian African American
Cases N = 2938 (%) Controls N = 3016 (%) Cases N = 1617 (%) Controls N = 1660 (%)
Age in years (mean and sd) 50 (8.5) 49 (8.3) 49 (8.1) 49 (8.1)
Income ($)
0–19,999 311 (11) 308 (10) 544 (35) 561 (35)
20,000–34,999 513 (18) 516 (18) 375 (24) 407 (25)
35,000–49,999 534 (19) 529 (18) 269 (17) 260 (16)
50,000–69,999 503 (18) 595 (20) 188 (12) 212 (13)
70,000 + 975 (33) 987 (34) 192 (12) 180 (11)
Site
Atlanta 559 (19) 551 (18) 321 (20) 343 (21)
Seattle 997 (34) 960 (32) 61 (4) 54 (3)
Detroit 400 (14) 446 (15) 278 (17) 333 (20)
Philadelphia 315 (11) 345 (11) 387 (24) 391 (24)
Los Angeles 667 (23) 714 (24) 570 (35) 539 (32)
Education
High school or less 948 (32) 1025 (34) 780 (48) 769 (46)
Tech school or some college 945 (32) 924 (31) 529 (33) 567 (34)
College graduate 1045 (36) 1067 (35) 307 (19) 324 (20)
Menopause status
Premenopausal 1368 (47) 1361 (45) 739 (46) 695 (42)
Postmenopausal 1273 (43) 1323 (44) 668 (41) 738 (44)
Unknown 297 (10) 332 (11) 210 (13) 227 (14)
Number of births
None 658 (22) 572 (19) 220 (14) 223 (13)
One 424 (14) 432 (14) 320 (20) 278 (17)
Two 957 (33) 922 (31) 396 (25) 415 (25)
Three 546 (19) 611 (20) 295 (18) 299 (18)
More than three 353 (12) 474 (16) 383 (24) 443 (27)
Oral contraceptive use
Never 652 (22) 608 (20) 386 (24) 382 (23)
Ever 2285 (78) 2406 (80) 1229 (76) 1276 (77)
Age at first birth (years)
<20 392 (13) 513 (17) 665 (41) 689 (42)
20–24 899 (31) 981 (33) 453 (28) 475 (29)
25–29 596 (20) 551 (18) 177 (11) 166 (10)
30 + 393 (13) 394 (13) 100 (6) 104 (6)
No births 658 (22) 572 (19) 220 (14) 223 (13)
Ever breastfed ‡2 weeks
Never 1616 (55) 1480 (49) 1071 (67) 1021 (62)
Ever 1320 (45) 1525 (51) 538 (33) 629 (38)
Age at menarche (years)
<12 731 (25) 777 (26) 459 (28) 482 (29)
12 1673 (57) 1674 (56) 845 (52) 790 (48)
‡13 526 (18) 561 (19) 311 (19) 386 (23)
HT use
Never 1661 (57) 1619 (54) 1166 (72) 1125 (68)
Estrogen + progestin 741 (25) 705 (23) 151 (9) 155 (9)
Estrogen only 449 (15) 600 (20) 256 (16) 332 (20)
160 Breast Cancer Res Treat (2008) 109:157–164
123
regardless of control group used. More recently, data from
two previously published Italian case-case control studies
were pooled along with an additional Swiss study [5].
Among the 6,315 cases and 6,038 controls, an OR of 0.72
(95% CI 0.62–0.85) was observed for a reported history of
ovarian cysts. The estimates of effect changed little
according to menopausal status, age at menarche, parity,
BMI, oophorectomy status, or other factors. The overall
results from these two most recent large pooled studies are
virtually identical. Previous studies have included only or
mostly Caucasian women. The present study is the first to
specifically report on the relationship between ovarian
cysts and breast cancer in African Americans.
Data from the Women’s CARE Study are consistent
with the two recent pooled analyses in that all the OR
estimates were less than 1.00. The results for the African
American women are quite similar to the recent pooled
analyses [4, 5]. Although the OR estimates for ovarian
cysts in Caucasians were also less than 1.0 in all subgroups,
the effect was not significant or was marginally significant
except for the significant OR of 0.85 (95% CI 0.76–0.96)
among all Caucasians combined. African Americans
overall had an OR of 0.68 (95% CI 0.57–0.81) for ovarian
cysts and the estimate was lower than the estimate in
Caucasians in all subgroups except for the small number of
women using progesterone only. As an ovarian cyst
Table 1 continued
Caucasian African American
Cases N = 2938 (%) Controls N = 3016 (%) Cases N = 1617 (%) Controls N = 1660 (%)
Progestestin only 85 (3) 88 (3) 43 (3) 48 (3)
Hysterectomy
Never 2318 (79) 2214 (73) 1129 (70) 1080 (65)
Ever 619 (21) 802 (27) 488 (30) 579 (35)
Number of ovaries removed
None 2559 (87) 2489 (83) 1329 (83) 1309 (79)
One 123 (4) 150 (5) 112 (7) 120 (7)
Two 255 (9) 373 (12) 168 (10) 224 (14)
BMI 5 years before reference date (kg/m2)
0–19.99 416 (14) 437 (15) 71 (4) 78 (5)
20.00–24.99 1519 (52) 1445 (48) 550 (34) 555 (34)
25.00–29.99 637 (22) 696 (23) 553 (34) 559 (34)
30.00 + 353 (12) 425 (14) 431 (27) 461 (28)
BMI at age 18 (kg/m2)
0–19.99 1581 (54) 1538 (51) 743 (46) 729 (44)
20.00–24.99 1188 (41) 1255 (42) 707 (44) 727 (44)
25.00 + 163 (6) 215 (7) 150 (9) 196 (12)
Smoking
Never 1334 (45) 1373 (46) 752 (46) 729 (44)
Former 1044 (36) 1012 (34) 445 (28) 454 (27)
Current 558 (19) 630 (21) 420 (26) 476 (29)
Alcohol use
Never 946 (32) 977 (32) 842 (52) 920 (55)
Former 506 (17) 546 (18) 313 (19) 277 (17)
Current 1482 (50) 1489 (49) 461 (29) 463 (28)
ER status
Positive 1907 (65) 733 (45)
Negative 733 (25) 578 (36)
Not available 298 (10) 306 (19)
PR status
Positive 1678 (57) 646 (40)
Negative 865 (29) 588 (36)
Not available 395 (13) 383 (24)
Breast Cancer Res Treat (2008) 109:157–164 161
123
Table 2 Odds ratios and 95% confidence intervals for the association between an ovarian cyst diagnosis and breast cancer incidence in all
Caucasian and African American women and selected subgroups in the CARE study
Caaucasian African American
Cases N (%) Controls N (%) ORa (95% CI) Cases N (%) Controls N (%) ORa (95% CI)
All
No cysts 2306 (78) 2281 (76) 1.00 1374 (85) 1315 (79) 1.00
Ovarian cysts 632 (22) 735 (24) 0.85 (0.76, 0.96) 243 (15) 345 (21) 0.68 (0.57, 0.81)
Premenopausal
No cysts 1109 (81) 1071 (80) 1.00 637 (86) 584 (84) 1.00
Ovarian cysts 259 (19) 278 (20) 0.91 (0.75, 1.10) 102 (14) 111 (16) 0.85 (0.63, 1.14)
Postmenopausal
No cysts 981 (77) 970 (73) 1.00 569 (85) 572 (77) 1.00
Ovarian cysts 292 (23) 353 (27) 0.86 (0.72, 1.03) 99 (15) 166 (23) 0.65 (0.49, 0.86)
Never used HT
No cysts 1353 (81) 1296 (80) 1.00 1009 (87) 926 (82) 1.00
Ovarian cysts 308 (19) 323 (20) 0.92 (0.78, 1.10) 157 (13) 199 (18) 0.73 (0.58, 0.91)
Estrogen + progestin
No cysts 589 (79) 551 (78) 1.00 124 (82) 120 (77) 1.00
Ovarian cysts 152 (21) 154 (22) 0.96 (0.75, 1.24) 27 (18) 35 (23) 0.78 (0.44, 1.37)
Estrogen only
No cysts 301 (67) 375 (62) 1.00 204 (80) 231 (70) 1.00
Ovarian cysts 148 (33) 225 (38) 0.82 (0.64, 1.07) 52 (20) 101 (30) 0.61 (0.42, 0.91)
Progestin only
No cysts 61 (72) 57 (65) 1.00 36 (84) 38 (79) 1.00
Ovarian cysts 24 (28) 31 (35) 0.67 (0.34, 1.33) 7 (16) 10 (21) 0.80 (0.26, 2.50)
No hysterectomy
No cysts 1901 (82) 1798 (81) 1.00 997 (88) 918 (85) 1.00
Ovarian cysts 417 (18) 416 (19) 0.97 (0.83, 1.13) 132 (12) 162 (15) 0.76 (0.59, 0.97)
Hysterectomy
No cysts 405 (65) 483 (60) 1.00 377 (77) 396 (68) 1.00
Ovarian cysts 214 (35) 319 (40) 0.80 (0.65, 1.00) 111 (23) 183 (32) 0.64 (0.49, 0.85)
No ovaries removed
No cysts 2134 (83) 2046 (82) 1.00 1186 (89) 1124 (86) 1.00
Ovarian cysts 425 (17) 443 (18) 0.94 (0.81, 1.09) 143 (11) 185 (14) 0.74 (0.58, 0.93)
One ovary removed
No cysts 88 (72) 109 (73) 1.00 67 (60) 64 (53) 1.00
Ovarian cysts 35 (28) 41 (27) 0.93 (0.54, 1.61) 45 (40) 56 (47) 0.55 (0.31, 0.96)
Both ovaries removed
No cysts 137 (54) 192 (51) 1.00 116 (69) 129 (58) 1.00
Ovarian cysts 118 (46) 181 (49) 0.93 (0.67, 1.29) 52 (31) 95 (42) 0.69 (0.44, 1.06)
Age at cyst diagnosis (years)
No cysts 2306 (79) 2281 (76) 1.00 1374 (85) 1315 (79) 1.00
<30 294 (10) 347 (12) 0.84 (0.71, 1.00) 95 (6) 142 (9) 0.64 (0.49, 0.84)
‡30 336 (11) 385 (13) 0.87 (0.74, 1.01) 147 (9) 202 (12) 0.70 (0.56, 0.88)
Time since cyst diagnosis (years)
No cysts 2306 (78) 2281(76) 1.00 1374 (85) 1325 (79) 1.00
<20 368 (13) 405 (13) 0.92 (0.79, 1.08) 159 (10) 215 (13) 0.71 (0.57, 0.89)
‡20 262 (9) 327 (11) 0.78 (0.65, 0.92) 83 (5) 129 (8) 0.61 (0.46, 0.82)
a Odds ratio and 95% confidence interval adjusted for age and study site
162 Breast Cancer Res Treat (2008) 109:157–164
123
diagnosis may be the cause or result of ovarian surgery, it
is difficult to separate any effects of the surgery or sub-
sequent menopausal status or use of HT from the effects of
the cysts. Although estimates varied slightly according to
menopausal, hormonal, or surgical subgroup, there was no
evidence for an interaction with any of these. Sampling
variation may be responsible for the variation in results
across different study populations. It is also interesting to
note, however, that in the Women’s CARE study, the racial
differences were restricted to ER and/or PR positive
tumours.
A biological explanation for an inverse relationship
between reported ovarian cysts and breast cancer is com-
plicated by the fact that ovarian cysts are heterogeneous and
the definition varies across studies. The term ‘‘ovarian
cysts’’ includes common functional cysts, which are most
frequently follicular cysts, but can also occur as luteal cysts
and theca lutein cysts [14]. Other types of cysts include
benign cystic teratomas, serous cystadenomas, and benign
serous tumours [14]. In addition, women with endometriotic
cysts or with PCOS may respond positively when queried
about a history of ovarian cysts. A major limitation in all
reports is their inability to distinguish among these condi-
tions. One exception is PCOS, which has been specifically
addressed in some studies. A report of increased risk asso-
ciated with PCOS in postmenopausal women in a cohort with
chronic anovulation [6] was followed by an inverse finding in
younger women [9] and two null reports [7, 8]. In general, the
prevalence of PCOS was quite low in these studies.
The literature on attributes associated with ovarian cysts
is sparse. Smoking has been associated with an increased
risk of unspecified ovarian cysts as well as with functional
cysts [15, 16], although BMI may modify the effect [16].
Functional cysts have also been associated with tubal
sterilization [11]. Endometriotic cysts have been associated
with regular periods, lower parity, and lower BMI while
serous and mucinous cysts have been associated with later
age at first birth, earlier menarche, and higher meat and
lower green vegetable consumption [10, 17]. The infor-
mation is currently insufficient to draw conclusions
with respect to a possible explanation for a relationship
with breast cancer. It is likely that the hypothalamic-
pituitary-gonadal axis plays a role in the development of
functional cysts, one of the more common types of ovarian
cysts, since it is responsible for follicular development. It
may be involved in other cyst types as well.
To confirm and make use of a finding of reduced breast
cancer risk associated with a diagnosis of ovarian cysts it
will be necessary to determine whether this reduced risk is
associated with one or more specific types of cyst, which
may act as a marker of the hormonal milieu. Such an
investigation would be possible in a large cohort of women
with detailed medical record information. This more spe-
cific knowledge would allow further research to determine
the protective mechanism.
Acknowledgements This study was funded by the National Insti-
tute of Child Health and Human Development, with additional sup-
port from the National Cancer Institute, through contracts with
Emory University (N01-HD-3-3168), Fred Hutchinson Cancer
Research Center (N01-HD-2-3166), Karmanos Cancer Institute at
Wayne State University (N01-HD-3-3174), University of Pennsyl-
vania (NO1-HD-3-3276), and University of Southern California
(N01-HD-3-3175) and through an interagency agreement with the
Centers for Disease Control and Prevention (Y01-HD-7022). The
Centers for Disease Control and Prevention contributed additional
staff and computer support. General support through SEER contracts
N01-PC-67006 (Atlanta), N01-CN-65064 (Detroit), N01-PC-67010
(Los Angeles), and N01-CN-0532 (Seattle) is also acknowledged.
Table 3 Odds ratios and 95% confidence intervals for the association between an ovarian cyst diagnosis and breast cancer incidence in
Caucasian and African American women according to the ER and PR status of the cases in the CARE study
Caucasian African American
Cases N (%) ORa (95% CI) Cases N (%) ORa (95% CI)
ER+
No cysts 1505 (79) 1.00 626 (85) 1.00
Ovarian cysts 402 (21) 0.84 (0.73, 0.97) 107 (15) 0.67 (0.53, 0.85)
ER-
No cysts 578 (79) 1.00 475 (82) 1.00
Ovarian cysts 155 (21) 0.84 (0.69, 1.03) 103 (18) 0.82 (0.64, 1.05)
PR+
No cysts 1333 (79) 1.00 552 (85) 1.00
Ovarian cysts 345 (21) 0.82 (0.70, 0.94) 94 (15) 0.66 (0.52, 0.85)
PR-
No cysts 672 (78) 1.00 478 (81) 1.00
Ovarian cysts 193 (22) 0.90 (0.75, 1.08) 110 (19) 0.87 (0.69, 1.11)
a Odds ratio and 95% confidence interval adjusted for age and study site
Breast Cancer Res Treat (2008) 109:157–164 163
123
The findings and conclusions in this report are those of the authors
and do not necessarily represent the views of the Centers for Disease
Control and Prevention.
References
1. Moseson M, Koenig KL, Shore RE, Pasternack BS (1993) The
influence of medical conditions associated with hormones on the
risk of breast cancer. Int J Epidemiol 22:1000–1009
2. Schairer C, Persson I, Falkeborn M, Naessen T, Troisi R, Brinton LA
(1997) Breast cancer risk associated with gynecologic surgery and
indications for such surgery. Int J Cancer 70:150–154
3. Weiss HA, Brinton LA, Potischman NA et al (1999) Breast
cancer risk in young women and history of selected medical
conditions. Int J Epidemiol 28:816–823
4. Knight JA, John EM, Milne RL et al (2006) Inverse association
between ovarian cysts and breast cancer in the breast cancer
family registry. Int J Cancer 118:197–202
5. Bosetti C, Scotti L, Negri E et al (2006) Benign ovarian cysts and
breast cancer risk. Int J Cancer 119:1679–1682
6. Coulam CB, Annegers JF, Kranz JS (1983) Chronic anovulation
syndrome and associated neoplasia. Obstet Gynecol 61:403–407
7. Anderson KE, Sellers TA, Chen P-L, Rich SS, Hong C-P, Folsom AR
(1997) Association of Stein-Leventhal syndrome with the incidence
of postmenopausal breast carcinoma in a large prospective study of
women in Iowa. Cancer 79:494–499
8. Parazzini F, La Vecchia C, Franceschi S, Talamini R, Negri E,
Crosignani PG (1997) Association of Stein-Leventhal syn-
drome with the incidence of postmenopausal breast carcinoma
in a large prospective study of women in Iowa. Cancer 80:
1360–1362
9. Gammon MD, Thompson WD (1991) Polycystic ovaries and the
risk of breast cancer. Am J Epidemiol 134:818–824
10. Parazzini F, La Vecchia C, Franceschi S, Negri E, Cecchetti G
(1989) Risk factors for endometrioid, mucinous and serous
benign ovarian cysts. Int J Epidemiol 18:108–112
11. Holt VL, Cushing-Haugen KL, Daling JR (2003) Oral contra-
ceptives, tubal sterilization, and functional ovarian cyst risk.
Obstet Gynecol 102:252–258
12. The ESHRE Capri Workshop Group (2001) Ovarian and endo-
metrial function during hormonal contraception. Hum Reprod
16:1527–1535
13. Marchbanks PA, McDonald JA, Wilson HG et al (2002) Oral
contraceptives and the risk of breast cancer. N Engl J Med
346:2025–2032
14. Berek JS (2002) Novak’s gynecology 13th edn. Lippincott
Williams and Wilkins, Philadelphia
15. Wyshak G, Frisch RE, Albright TE, Albright NL, Schiff I (1988)
Smoking and cysts of the ovary. Int J Fertil 33:398–404
16. Holt VL, Cushing-Haugen KL, Daling JR (2005) Risk of func-
tional ovarian cyst: effects of smoking and marijuana use
according to body mass index. Am J Epidemiol 161:520–525
17. Chiaffarino F, Parazzini F, Surace M, Benzi G, Chiantera V, La
Vecchia C (2003) Diet and risk of seromunicous benign ovarian
cysts. Eur J Obstet Gynecol Reprod Biol 2003;110:196–200
164 Breast Cancer Res Treat (2008) 109:157–164
123