University of Southern California Dissertations and Theses

The effects of hormonal exposures on ovarian and breast cancer risk

(USC Thesis Other)

The effects of hormonal exposures on ovarian and breast cancer risk

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

THE EFFECTS OF HORMONAL EXPOSURES ON OVARIAN AND BREAST
CANCER RISK

by
Alice Wen-Ron Lee

________________________________________________________________________

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(EPIDEMIOLOGY)

May 2016

Copyright 2016 Alice Wen-Ron Lee
ii

Acknowledgements
I would like to express my deepest appreciation to Dr. Celeste Leigh Pearce who
has provided me with both invaluable knowledge and opportunities as I pursued my
graduate studies. She has been a mentor, a role model, and a friend, and I am truly
grateful for her mentorship over the years. In addition, I would like to take this
opportunity to thank my other committee members, Drs. Anna Wu, Lynda Roman, and
Daniel Stram, for their time and their guidance. They have been instrumental in my
training as an epidemiologist.
Lastly, I would also like to thank my family and friends whose support and
encouragement have meant the world to me.

iii

Table of Contents

Acknowledgements ............................................................................................................. ii

List of Tables .......................................................................................................................v

List of Figures ................................................................................................................... vii
Abstract ............................................................................................................................ viii
Chapter 1: Introduction ........................................................................................................1
Chapter 2: Background on Ovarian Cancer .......................................................................12
Trends and Statistics ..............................................................................................14
Tumor Histotype and Classification ......................................................................15
Clinical Features: Screening, Symptoms, and Treatment ......................................18
Genetic and Environmental (“Non-Genetic”) Risk Factors ...................................23
References ..............................................................................................................35

Chapter 3: Background on Hormone Therapy ...................................................................43
Overview of Hormone Therapy .............................................................................45
Critical Evaluation of the Hormone Therapy-Ovarian Cancer Literature .............53
References ..............................................................................................................67

Chapter 4: Association Between Menopausal Estrogen-Only Therapy and Ovarian
Carcinoma Risk ..................................................................................................................77
Materials and Methods ...........................................................................................84
Results ....................................................................................................................88
Discussion ..............................................................................................................90
References ............................................................................................................102
Chapter 5: The Association Between Menopausal Estrogen-Progestin Therapy and
Risk of Ovarian Cancer: A Pooled Analysis....................................................................107
Materials and Methods .........................................................................................110
Results ..................................................................................................................115
Discussion ............................................................................................................117
References ............................................................................................................130
Chapter 6: A Splicing Variant of TERT Identified by GWAS Interacts with
Menopausal Estrogen Therapy in Risk of Ovarian Cancer ............................................136
Materials and Methods .........................................................................................140
Results ..................................................................................................................144
Discussion ............................................................................................................146
References ............................................................................................................157
iv

Chapter 7: The Effects of New Generation Hormonal Birth Control Methods on
Breast Cancer Risk Among Young Women: A Population-based Study ........................163
Specific Aims .......................................................................................................166
Research Strategy.................................................................................................169
References ............................................................................................................188

Chapter 8: Conclusion......................................................................................................191
Summary of Chapters 4-6 ....................................................................................194
Summary of Chapter 7 .........................................................................................199
Concluding Remarks ............................................................................................200
References ............................................................................................................202
v

List of Tables

Table 2.1. Three Main Staging Systems for Ovarian Cancer ............................................28

Table 2.2. Overall FIGO Stage Distribution and 5-year Relative Survival by Stage at
Diagnosis for Ovarian Cancer, 2000-2012, All Races Combined .....................................29

Table 2.3. Common Grading Systems for Ovarian Cancer ...............................................30

Table 2.4. Distribution and Characteristics of the Main Ovarian Cancer Histotypes ........31

Table 2.5. Five-year Ovarian Cancer Survival Rates by Histotype and FIGO Stage at
Diagnosis ...........................................................................................................................32

Table 2.6. Type of Association Between Each Etiologic Factor and Ovarian Cancer
Risk by Histotype Based on Large Pooled Analyses .........................................................33

Table 3.1. Progestin Doses by EPT Regimen in the United States and Sweden ...............60

Table 3.2. Current Literature on the Association of ET Use on Risk of Serous and
Mucinous Ovarian Cancer .................................................................................................61

Table 3.3. Current Literature on the Association of ET Use on Risk of Endometrioid
and Clear Cell Ovarian Cancer ..........................................................................................62

Table 3.4. Current Literature on the Association Between EPT Use and Risk of
Ovarian Cancer by EPT Regimen ......................................................................................63

Table 4.1. Characteristics of Studies Included in Analysis ...............................................93

Table 4.2. Numbers of Cases and Controls Included in Analysis by Study ......................95

Table 4.3. Hormone Therapy Questionnaire Questions by Study .....................................96

Table 4.4. Association Between ET Use Over Age 50 and Risk of Ovarian Carcinoma
Overall................................................................................................................................97

Table 4.5. Association Between ET Use After Age 50 and Risk of Serous and
Endometrioid Ovarian Carcinoma .....................................................................................98

vi

Table 4.6. Association Between ET Use After Age 50 and Risk of Mucinous and
Clear Cell Ovarian Carcinoma ...........................................................................................99

Table 4.7. Association Between Current-or-Recent ET use and Risk of Ovarian
Carcinoma by Histotype Using Age at Hysterectomy as Age at Menopause..................100

Table 5.1. Numbers of Cases and Controls Included in Analysis by Study ....................122

Table 5.2. Hormone Therapy Questionnaire Questions by Study ...................................123

Table 5.3. Association Between EPT Use and Risk of Ovarian Cancer Overall .............125

Table 5.4. Association Between EPT Use and Risk of Ovarian Cancer by Histotype ....126

Table 5.5. Association Between sEPT Use and cEPT Use and Risk of Ovarian Cancer
by Histotype .....................................................................................................................127

Table 5.6. Association Between sEPT Use and cEPT Use and Risk of Ovarian Cancer
Overall by Duration .........................................................................................................128

Table 6.1. Description of Studies Included in Analysis ..................................................150

Table 6.2. Characteristics of the 18 SNPs Included in the Analysis and Their
Previously Reported Best Hits .........................................................................................151

Table 6.3. Association Between the 18 SNPs and Serous Ovarian Cancer Risk.............152

Table 6.4. Association Between the 18 SNPs and Risk Score and Serous Ovarian
Cancer Risk, Stratified by ET Use After Age 50 .............................................................153

Table 6.5. Association Between the 18 SNPs and Endometrioid Ovarian Cancer Risk .154

Table 6.6. Association Between the 18 SNPs and Risk Score and Endometrioid
Ovarian Cancer Risk, Stratified by ET Use After Age 50 ...............................................155

Table 6.7. Association Between Rs10069690 and Serous Ovarian Cancer Risk
Among ET Users, Stratified by Duration of ET Use After Age 50 .................................156

Table 7.1. Sample Size Needed by Risk Estimate With a Power=80% and a Two-sided
Alpha=0.05 ......................................................................................................................186
vii

List of Figures

Figure 2.1. Age-adjusted Incidence and Mortality Rates for Ovarian Cancer,
2000 to 2012, All Races Combined ...................................................................................34

Figure 3.1. Total Number of Adult Women Using One or More Prescribed Hormone
Therapy Drugs in the Non-Institutionalized Population of the United States, 2001 to
2008....................................................................................................................................64

Figure 3.2. Forest Plot of Study-specific and Summary RRs and 95% CIs for Risk of
Ovarian Cancer Per 5 Years of EPT Use from Pearce et al's Meta-analysis .....................65

Figure 3.3. Forest Plot of Study-specific and Summary RRs and 95% CIs for Risk of
Ovarian Cancer Per 5 Years of EPT Use from Pearce et al's Meta-analysis .....................66

Figure 4.1. Flowchart of Analysis Exclusions .................................................................101

Figure 5.1. Flowchart of Analysis Exclusions .................................................................129

Figure 7.1. Age-incidence Curve of Breast Cancer Based on Data from White
Women in the United States, 1969-1971 .........................................................................187

viii

Abstract
Ovarian cancer accounts for more deaths than any other cancer in the female
reproductive system. With a poor survival rate due to a lack of a routine screening
method, elucidating its etiology is imperative for disease prevention. The current
literature suggests ovarian cancer’s etiology to be hormone-related as exemplified by
some of its well-established etiologic factors, such as oral contraceptive use and parity,
which affect endogenous levels of estrogen and progesterone. Hormone therapy use also
affects these hormone levels, but its impact on ovarian cancer risk remains to be clarified.
This dissertation comprehensively examines the hormone therapy-ovarian cancer
association, considering the type used, the duration and recency of its use, and disease
histotype. Given that estrogen and progesterone are hypothesized to have distinct effects
on the ovary and hormone therapy entails synthetic forms of these hormones, a better
understanding of this association can have significant clinical implications and contribute
to our understanding of ovarian carcinogenesis. This dissertation also explores potential
gene-environment interactions in the context of hormone therapy, which may be useful
for future lifetime risk estimation.
Lastly, oral contraceptive use has been shown to increase risk of breast cancer,
but how this increased risk is impacted by new birth control methods and regimens
remains uncertain. This dissertation concludes with an epidemiologic study proposal to
address this research question in order to evaluate the effects of another exogenous
hormonal exposure on a malignancy that is often studied with ovarian cancer.

Chapter 1: Introduction

1

Abbreviation List:
EPT – estrogen-progestin combined therapy
ET – estrogen-only therapy
GWAS – genome-wide association study
HT – hormone therapy
OC – oral contraceptive
OCAC – Ovarian Cancer Association Consortium

Running Head:
Introduction

2

This dissertation is written in accordance with the requirements of the Doctor of
Philosophy (Ph.D.) degree in Epidemiology in the Department of Preventive Medicine at
the University of Southern California Keck School of Medicine. The purpose of this
dissertation is to explore the effects of hormonal exposures on ovarian and breast cancer
risk. This is accomplished by: 1) a background on ovarian cancer, 2) an overview of
hormone therapy (HT) and a critical review of the current literature on its relationship
with ovarian cancer, 3) three independent analyses that comprehensively evaluate this
association (presented as three manuscripts), and 4) a grant proposal to explore the effects
of new generation oral contraceptives (OCs) and other hormonal birth control options on
breast cancer risk among young women.

Ovarian cancer is the leading cause of mortality among all gynecologic cancers in
the United States (1). This is in large part due to the difficulty in detecting early-stage
disease since there is currently no effective screening method. In September 2012, the
United States Preventive Services Task Force recommended against screening for ovarian
cancer in women who do not exhibit clinical signs and symptoms since routine screening
had shown no proven benefit and may lead to serious harms, such as major surgical
interventions among those who do not have the disease (2). While OCs are an effective
chemopreventive strategy, a better understanding of the etiology of ovarian cancer is
sorely needed.
Ovarian cancer’s etiology is likely to be hormone-related. This is exemplified by
several well-established risk and protective factors including OC use, parity, and having a
personal history of endometriosis (3). Given that these factors are directly and indirectly
3

related to estrogen and progesterone levels, it is logical to evaluate how HT, which
consists of synthetic forms of these hormones, affects ovarian cancer risk. Most studies
have found an increased risk of ovarian cancer with HT use, but there are features to this
association that remain to be clarified, including the effects of different HT types
(estrogen-only therapy [ET] versus estrogen-progestin combined therapy [EPT]) and EPT
regimens (sequential EPT versus continuous EPT), duration and recency of HT use as
well as disease histotype. These aspects have been examined in various individual
studies, but with inconsistent results partly due to inadequate power. Most recently, the
Collaborative Group on Epidemiological Studies on Ovarian Cancer (Collaborative
Group) conducted a pooled analysis of 52 epidemiological studies to assess some of these
features with a greater sample size (4). However, some of their findings were contrary to
a previous comprehensive meta-analysis that included many of the same studies (5).
Ovarian cancer is also known to have a strong genetic component, which is partly
attributable to both high-penetrance susceptibility mutations as well as common variants
identified using genome-wide association studies (GWASs). There are currently 18
confirmed susceptibility loci, each associated with a modest relative risk estimate (6-12).
However, it is possible that interactions between non-genetic etiologic factors, such as
HT, and genetic etiologic factors, such as these confirmed loci, exist thereby putting
some women at higher risk.
To better understand the HT-ovarian cancer association and address some of these
inconsistencies and gaps in the current literature, data from the Ovarian Cancer
Association Consortium (OCAC)
(http://apps.ccge.medschl.cam.ac.uk/consortia/ocac/index.html), an international
4

collaboration of ovarian cancer studies formed in 2005 with the goal of identifying genes
and risk factors related to ovarian cancer risk, were used in this dissertation. Briefly,
primary data were sent by each study investigator to the consortium data coordinating
center at Duke University (Durham, North Carolina). The epidemiology working group
of the OCAC cleaned these data and resolved any inconsistencies to generate a core
dataset that included self-reported demographics, tumor characteristics, and basic ovarian
cancer risk and protective factor information. Additional HT data on the specific EPT
regimens that were not part of the core dataset were requested from studies whose
questionnaires indicated appropriate collection of this information. The genetic data were
based on three GWASs, their replication efforts, and two large-scale genotyping arrays,
which were combined with data from the 1000 Genomes Project so imputation could be
carried out for all OCAC participants. In total, 11 population-based case-control studies
participating in the OCAC were included in this dissertation; each manuscript was based
on a subset of these eleven studies (13-23).
The first part of this dissertation (Chapters 2 and 3) provides a clinical and
epidemiologic background on ovarian cancer, setting the stage for assessing the effect of
HT on disease risk. It also includes an overview of HT, including a discussion of its
various types and the biological and epidemiologic evidence supporting its role in ovarian
carcinogenesis, as well as a critical evaluation of the current literature on HT’s
association with ovarian cancer.
The second part of this dissertation (Chapters 4 through 6) consists of the three
manuscripts. The first manuscript (Chapter 4) comprehensively assesses the association
between ET use and risk of ovarian cancer, considering duration and recency of use as
5

well as histotype-specific effects. It focuses specifically on hysterectomized women since
they constitute the population most likely to use ET since ET is a confirmed risk factor
for endometrial cancer (24). The second manuscript (Chapter 5) evaluates the effects of
EPT use on risk of ovarian cancer, taking a similar approach as the first manuscript, but
among all women since its use is not contraindicated for a subset of the population. It
also assesses the effects of the two EPT regimens to better understand the role of
progesterone. The third manuscript (Chapter 6) explores potential gene-environment
interactions between ET use and the 18 confirmed ovarian cancer common variants on
disease risk.
The third part of this dissertation is a grant proposal for a population-based case-
control study to assess the effects of Depo Provera, Mirena, third and fourth generation
progestin OCs, and other new OC regimens, on breast cancer risk among young women.
Breast cancer is the most common malignancy among American women, including those
under the age of 40 (25). Many studies have shown that use of OCs is associated with an
increased risk of breast cancer, but that this risk diminishes with time and ceases 10 years
after stopping use (26). However, birth control formulations have significantly changed
over the years and how these new options may impact disease risk is not well-understood.
These are critical questions that need to be addressed, especially given the growing
popularity of these new birth control options. In addition, the study proposes to
implement a novel control recruitment strategy using social media. Traditional
recruitment methods have become increasingly problematic and often face low
participation rates. This proposal includes a pilot study to first test the feasibility of using
social media for control recruitment with the hopes of implementing this new strategy in
6

the proposed study. If this method is unsuccessful, traditional recruitment methods, such
as neighborhood controls or random digit dialing, will be used instead.
Hormones play important roles in the development of ovarian and breast cancer.
In order to successfully create more effective prevention strategies, these roles need to be
clarified. As we uncover the genetic and non-genetic aspects that contribute to their risk,
we move steps closer to elucidating their etiologies, which will improve early detection
and future prevention.

7

REFERENCES
1. American Cancer Society. Ovarian Cancer.
(http://www.cancer.org/acs/groups/cid/documents/webcontent/003130-pdf.pdf).
(Accessed February 2, 2016).
2. Moyer VA, U.S Preventive Services Task Force. Screening for ovarian cancer:
U.S. Preventive Services Task Force reaffirmation recommendation statement.
Ann Intern Med. 2012;157(12):900-904.
3. Pearce CL, Rossing MA, Lee AW, et al. Combined and interactive effects of
environmental and GWAS-identified risk factors in ovarian cancer. Cancer
Epidemiol Biomarkers Prev. 2013;22(5):880-890.
4. Collaborative Group On Epidemiological Studies Of Ovarian C, Beral V,
Gaitskell K, et al. Menopausal hormone use and ovarian cancer risk: individual
participant meta-analysis of 52 epidemiological studies. Lancet.
2015;385(9980):1835-1842.
5. Pearce CL, Chung K, Pike MC, et al. Increased ovarian cancer risk associated
with menopausal estrogen therapy is reduced by adding a progestin. Cancer.
2009;115(3):531-539.
6. Bolton KL, Tyrer J, Song H, et al. Common variants at 19p13 are associated with
susceptibility to ovarian cancer. Nat Genet. 2010;42(10):880-884.
7. Goode EL, Chenevix-Trench G, Song H, et al. A genome-wide association study
identifies susceptibility loci for ovarian cancer at 2q31 and 8q24. Nat Genet.
2010;42(10):874-879.
8

8. Song H, Ramus SJ, Tyrer J, et al. A genome-wide association study identifies a
new ovarian cancer susceptibility locus on 9p22.2. Nat Genet. 2009;41(9):996-
1000.
9. Pharoah PD, Tsai YY, Ramus SJ, et al. GWAS meta-analysis and replication
identifies three new susceptibility loci for ovarian cancer. Nat Genet.
2013;45(4):362-370.
10. Shen H, Fridley BL, Song H, et al. Epigenetic analysis leads to identification of
HNF1B as a subtype-specific susceptibility gene for ovarian cancer. Nat
Commun. 2013;4:1628.
11. Bojesen SE, Pooley KA, Johnatty SE, et al. Multiple independent variants at the
TERT locus are associated with telomere length and risks of breast and ovarian
cancer. Nat Genet. 2013;45(4):371-384.
12. Kuchenbaecker KB, Ramus SJ, Tyrer J, et al. Identification of six new
susceptibility loci for invasive epithelial ovarian cancer. Nat Genet.
2015;47(2):164-171.
13. Wu AH, Pearce CL, Tseng CC, et al. African Americans and Hispanics Remain at
Lower Risk of Ovarian Cancer Than Non-Hispanic Whites after Considering
Nongenetic Risk Factors and Oophorectomy Rates. Cancer Epidemiol
Biomarkers Prev. 2015;24(7):1094-1100.
14. Risch HA, Bale AE, Beck PA, et al. PGR +331 A/G and increased risk of
epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2006;15(9):1738-
1741.
9

15. Balogun N, Gentry-Maharaj A, Wozniak EL, et al. Recruitment of newly
diagnosed ovarian cancer patients proved challenging in a multicentre biobanking
study. J Clin Epidemiol. 2011;64(5):525-530.
16. Glud E, Kjaer SK, Thomsen BL, et al. Hormone therapy and the impact of
estrogen intake on the risk of ovarian cancer. Arch Intern Med.
2004;164(20):2253-2259.
17. Terry KL, De Vivo I, Titus-Ernstoff L, et al. Androgen receptor cytosine, adenine,
guanine repeats, and haplotypes in relation to ovarian cancer risk. Cancer Res.
2005;65(13):5974-5981.
18. Ness RB, Dodge RC, Edwards RP, et al. Contraception methods, beyond oral
contraceptives and tubal ligation, and risk of ovarian cancer. Ann Epidemiol.
2011;21(3):188-196.
19. Moorman PG, Calingaert B, Palmieri RT, et al. Hormonal risk factors for ovarian
cancer in premenopausal and postmenopausal women. Am J Epidemiol.
2008;167(9):1059-1069.
20. Bodelon C, Cushing-Haugen KL, Wicklund KG, et al. Sun exposure and risk of
epithelial ovarian cancer. Cancer Causes Control. 2012;23(12):1985-1994.
21. Lurie G, Terry KL, Wilkens LR, et al. Pooled analysis of the association of
PTGS2 rs5275 polymorphism and NSAID use with invasive ovarian carcinoma
risk. Cancer Causes Control. 2010;21(10):1731-1741.
22. Bandera EV, King M, Chandran U, et al. Phytoestrogen consumption from foods
and supplements and epithelial ovarian cancer risk: a population-based case
control study. BMC Womens Health. 2011;11:40.
10

23. Royar J, Becher H, Chang-Claude J. Low-dose oral contraceptives: protective
effect on ovarian cancer risk. Int J Cancer. 2001;95(6):370-374.
24. Beral V, Bull D, Reeves G, et al. Endometrial cancer and hormone-replacement
therapy in the Million Women Study. Lancet. 2005;365(9470):1543-1551.
25. Johnson RH, Chien FL, Bleyer A. Incidence of breast cancer with distant
involvement among women in the United States, 1976 to 2009. JAMA.
2013;309(8):800-805.
26. Breast cancer and hormonal contraceptives: collaborative reanalysis of individual
data on 53 297 women with breast cancer and 100 239 women without breast
cancer from 54 epidemiological studies. Collaborative Group on Hormonal
Factors in Breast Cancer. Lancet. 1996;347(9017):1713-1727.

11
Chapter 2: Background on Ovarian Cancer

12
Abbreviation List:
AJCC/TNM – American Joint Committee on Cancer/Tumor, Node, Metastasis
BMI – body mass index
CA-125 – cancer antigen 125
FIGO – International Federation of Gynecology and Obstetrics
OC – oral contraceptive
RR-BSO – risk-reducing bilateral salpingo-oophorectomy
SEER – Surveillance, Epidemiology, and End Results
TVU – transvaginal ultrasound
U.S. – United States

Running Head:
Background on Ovarian Cancer

13
Ovarian cancer accounts for more deaths than any other cancer in the female
reproductive system (1). In addition, it is associated with a significant burden of disease
for both the individual and society. As the majority of ovarian cancer patients are
diagnosed at late stages when survival is poor, it is imperative that we better understand
its etiology so improved prevention strategies can be developed.

TRENDS AND STATISTICS
Approximately 21,290 women in the United States (U.S.) are projected to receive
a new diagnosis of ovarian cancer and 14,180 to die from ovarian cancer in 2015 (1). The
average lifetime risk of the disease for women in the U.S. is 1.3%, which is low relative
to the risk of cancer in other sites, such as breast cancer which has a lifetime risk of 12%
(1, 2). However, ovarian cancer is the ninth most common cancer among women in the
U.S., excluding non-melanoma skin cancer, and ranks fifth in female cancer deaths (1).
According to data from registries included in Surveillance, Epidemiology, and
End Results (SEER) 9, the annual age-adjusted incidence rate in 2012 was 11.9 cases per
100,000 women while the annual age-adjusted death rate was 7.4 deaths per 100,000
women for all races combined (3). As shown in Figure 2.1, incidence and mortality have
decreased 16.8% and 16.9%, respectively, from 2000 to 2012. These decreasing trends
are even more striking when earlier years are considered; the annual age-adjusted
incidence rate in 1975 was 16.3 cases per 100,000 women whereas the annual age-
adjusted death rate was 9.8 deaths per 100,000 women.
When these rates are looked at by histotype from 2000 to 2012, the decreasing
trend is most noticeable for the high-grade serous and endometrioid histotype for both
14
incidence and mortality. The rates for the mucinous histotype also showed slight
decreases, but remained relatively unchanged for the clear cell and low-grade serous
histotypes.

TUMOR HISTOTYPE AND CLASSIFICATION
There are 3 main types of ovarian tumors depending on the cells from which the
tumor develops: germ cell, stromal, and epithelial. Epithelial tumors constitute about 85%
to 90% of all ovarian cancers (1). They can be classified as borderline (low-malignant
potential), which are slower-growing with no evidence of stromal invasion, or invasive
(malignant); only 15% of epithelial ovarian cancers are diagnosed as borderline (4, 5).
The focus of this dissertation will be on invasive, epithelial tumors, which have been and
will be referred to as ovarian cancers herein.
The stage of ovarian cancer refers to the cancer's spread and can only be
determined after a thorough examination of tissue samples obtained during surgery. It is
directly correlated with patient survival as the later the stage of diagnosis, the poorer the
survival. Stage can also affect the patient’s treatment course as some who are diagnosed
with early-stage disease may only require surgery since the benefits of chemotherapy for
such patients are uncertain (6). One study found lack of proper surgical staging to be an
important independent factor in predicting future recurrence (7). The Gynecologic
Oncology Group's definition of complete surgical staging includes the following: 1) total
hysterectomy with bilateral salpingo-oophorectomy; 2) infracolic omentectomy; 3)
aspiration of free peritoneal fluid; 4) peritoneal washings for cytology (abdomen and
pelvis); 5) inspection of all abdominal peritoneal surfaces; 6) peritoneal biopsies from
15
four pelvic locations and bilateral paracolic areas; 7) diaphragm scraping or biopsy; and
8) bilateral selective pelvic and aortic lymph node dissections (6).
There are three main staging systems for ovarian cancer: the American Joint
Committee on Cancer/Tumor, Node, Metastasis (AJCC/TNM), the International
Federation of Gynecology and Obstetrics (FIGO), and SEER. The AJCC/TNM describes
stage in terms of the tumor's extent (T), whether it has spread to nearby lymph nodes (N),
and whether it has metastasized (M). Various combinations of T, N, and M correspond to
the FIGO stages (stages I to IV) with the AJCC/TNM and FIGO staging systems closely
resembling one another. The three main SEER stages describe the overall spread of the
cancer as localized, regional, or distant. The relationships among these three systems are
presented in Table 2.1. Table 2.2 shows the FIGO stage distribution of ovarian cancer
patients as well as each stage's 5-year relative survival using SEER (8). It clearly
demonstrates ovarian cancer’s fatality as more than half of the patients are diagnosed in
stages III and IV.
Ovarian cancer is further classified by grade, which is based on how "normal" the
tumor tissue appears under the microscope. Grade is an indicator of how likely the tumor
will grow and spread and like stage, it can impact a patient’s treatment course. One of the
most commonly used grading systems is the FIGO system, which uses a three-tier
scheme (low-, intermediate-, or high-grade) to reflect the level of cellular organization
into differentiated structures such as glands and papillae (9). In addition, Malpica et al at
the M.D. Anderson Cancer Center proposed a two-tier system (low- or high-grade) for
serous ovarian cancers, which is based on nuclear atypia and the mitotic rate (10).
Generally, low-grade tumors develop slowly from well-recognized precursors whereas
16
high-grade tumors evolve rapidly and are more likely to spread (11). If a grading system
is not specified, a tumor's grade is typically based on a scale from one to four, with grade
1 indicating well-differentiated cancer cells and grade 4 indicating undifferentiated
cancer cells (12). Table 2.3 summarizes these various grading systems. The studies
included in this dissertation graded tumors using the four-level scale. A two-tier scheme
will be applied to this scale in which those classified as grade 1 will be considered low-
grade and those classified as grades 2 to 4 will be considered high-grade.
Lastly, ovarian cancers can be classified by histotype with the main types being
high-grade serous, low-grade serous, mucinous, endometrioid, and clear cell. Each not
only varies in frequency, with high-grade serous constituting the majority of ovarian
cancers, but is also characterized by various pathologic, genetic, and epidemiologic
features (13, 14). In addition, each histotype has a different hypothesized cell of origin
and unique molecular and genetic alterations associated with it (13). The most prevalent
current view is that serous tumors develop from the fimbriated portion of the fallopian
tube, endometrioid and clear cell tumors from endometriosis, and mucinous tumors from
transitional-type epithelium located at the tubal-mesothelial junction or are misdiagnosed
colon cancers (15). Table 2.4 describes each histotype’s distinct characteristics.
The distinction between low- and high-grade has played an important role in the
characterization of disease histotype. A dualistic model was proposed that categorizes
various types of ovarian cancers into two groups: type I tumors, which are clinically
indolent and include low-grade serous, low-grade endometrioid, clear cell, and mucinous
cancers, and type II tumors, which are more aggressive and include high-grade serous
and high-grade endometrioid cancers (15). Low- and high-grade serous ovarian cancers
17
have recently been recognized as two distinct histotypes due to differing morphologic
features as well as the rareness of low-grade serous tumors progressing to higher-grade
tumors (13).

CLINICAL FEATURES: SCREENING, SYMPTOMS, TREATMENT
Screening
Generally, a disease that is appropriate for routine screening must be an important
cause of mortality, be significantly prevalent, have a detectable preclinical phase, and be
amenable to therapy (16). The first and last conditions are met for ovarian cancer, but
whether it has a detectable preclinical phase is unclear. Although not widely accepted,
Brown et al recently developed mathematical models to describe the early natural history
of ovarian cancer (17). They estimated that serous tumors spend approximately 4.3 years
as histopathologically detectable, which could represent this necessary "window" (17).
In addition, while ovarian cancer is relatively rare, it is significantly prevalent among
women who carry high-risk mutations in major ovarian cancer genes (see “Genetic
factors” section below), suggesting that this population would be appropriate for
screening.
One screening study that focuses on women in high-risk populations is the
ongoing United Kingdom Familial Ovarian Cancer Screening Study, which includes
women who have a minimum estimated lifetime ovarian cancer risk of 10% on the basis
of family history or predisposing mutations (18). This screening study has no random
assignment to a nonscreening arm since this was deemed unacceptable to high-risk
women and clinicians. The purpose of Phase I of this study was to establish the
18
performance characteristics of annual transvaginal ultrasound (TVU) and serum cancer
antigen 125 (CA-125) screening and to investigate the impact of delayed screening
interval and surgical intervention; CA-125 is a protein found on the surface of many
ovarian cancer cells, which is why it is often used as a tumor marker. The results were
encouraging as such a screening method had a sensitivity of greater than 80% and
positive and negative predictive values of 25.5% and 99.9%, respectively. More
importantly, those who were not screened the year before were significantly more likely
to have advanced-stage disease, emphasizing the importance of rigorous adherence to
screening schedules among those characterized as high-risk. Phase II will consist of
screening every four months and is currently in follow-up.
As mentioned previously, given ovarian cancer's rareness and lack of a detectable
preclinical phase, it does not meet the criteria for routine screening. Ovarian cancer is
uncommon among the general population and Brown et al's model describing the early
natural history of the disease is not widely accepted. However, its fatality has prompted
several randomized controlled trials to evaluate the efficacy of TVU and elevated CA-
125 levels as routine screening tools. The Prostate, Lung, Colorectal and Ovarian cancer
screening trial randomized women in the general population to compare those receiving
annual screening with CA-125 for six years and TVU for four years to those receiving
their usual medical care (ie. not offered annual screening with CA-125 or TVU) (19).
They found no difference in the stage at diagnosis and the mortality rate between the two
groups, but did find a significantly higher rate of complications due to further diagnostic
evaluations from false-positive results. Thus, the U.S. Preventive Services Task Force
19
concluded that the harms outweigh the benefits when using these screening methods
routinely (20).

Symptoms
Symptom-based screening has also been considered, although its lack of
specificity and the rareness of ovarian cancer raise concerns with regard to its accuracy
and the healthcare costs associated with using such a method. While there have been no
symptom-based screening trials to date, pilot studies have found that women with non-
specific symptoms were willing to undergo such screening, and that this type of screening
was feasible and acceptable in a clinical setting (21, 22). In 2007, the Gynecologic
Cancer Foundation, the Society for Gynecologic Oncologists, and the American Cancer
Society released a consensus statement indicating that certain symptoms, including
bloating, abdominal and pelvic pain, difficulty eating or feeling full quickly, and urinary
urgency or frequency, are more common among ovarian cancer patients than women in
the general population (23). They also recommended that women consult a physician if
they experience these symptoms daily as it could lead to earlier diagnoses (23).
This consensus statement was partly motivated by Goff et al's study, which had
developed a symptom index based on the frequency, severity, and duration of the
symptoms (24). Such an evaluation had been effective in identifying women with ovarian
cancer in their study and may be a possible way to detect disease earlier (24). However,
Rossing et al's study assessed the sensitivity, specificity, and positive predictive value of
both Goff's symptom index as well as of the symptoms included in the consensus
statement and found that while the sensitivities ranged from 53.2% to 73.4% and the
20
specificities were all above 92%, the positive predictive values were no higher than 1.1%
(25); the use of symptoms to indicate evaluation for disease would only result in the
diagnosis of one out of 100 women in the general population exhibiting such symptoms.

Treatment
Surgery is the cornerstone for the treatment of ovarian cancer. It is first used for
diagnosis, then staging, followed by cytoreduction for patients with advanced-stage
disease. Cytoreductive or debulking surgery entails the surgical removal of the tumor
mass is considered to be one of the most important prognostic factors for survival (26). A
review of 11 retrospective studies on cytoreductive surgery among women with
advanced-stage disease concluded that complete cytoreduction, resulting in microscopic
or no visible disease, was strongly associated with overall and progression-free survival
(27). Complete cytoreduction is the surgical goal, but when this is not achievable, optimal
cytoreduction, which leaves less than 1 cm of residual disease, is the goal (26).
Chemotherapy plays a key role in disease management. It includes a combination
of a platinum compound, such as cisplatin or carboplatin, which damages the DNA and
limits cell division, and a taxane, such as paclitaxel or docetaxel, which halts mitosis and
causes apoptosis. Each patient requires an individualized treatment course depending on
the stage and grade of her disease, although the standard, preferred regimen for patients
with advanced disease is six cycles of carboplatin and paclitaxel (28). For advanced
tumors, chemotherapy will typically follow optimal surgical cytoreduction, which should
achieve a complete remission for most patients. However, approximately 70% of
advanced cases relapse, and even in stage I or II patients, the relapse rate is 20% to 25%,
21
with tumors becoming increasingly chemoresistant (28, 29). Recently, there has been
growing interest in neoadjuvant chemotherapy, which entails delivering chemotherapy
prior to a cytoreductive operation. The motivation behind this is to first reduce tumor
burden and improve patient performance status since aggressive surgical cytoreduction
can be detrimental. However, studies have shown that patients undergoing neoadjuvant
chemotherapy had similar survival compared to those undergoing debulking surgery prior
to chemotherapy (30, 31). There is currently no distinct standard protocol for the use of
neoadjuvant chemotherapy defined or accepted by clinicians (28).
Chemotherapy's overall impact on stage I disease is debatable since clinical trials
have had limited power to assess its effects on cancer recurrence or survival. While it has
been shown to be associated with improved recurrence-free survival in early-stage
patients, it appears to be limited to only those who had incomplete surgical staging,
which calls into question the study participants' true tumor stage since thorough surgical
staging procedures have been shown to upstage the diagnosis in 10-30% of cases (28,
32). The National Institutes of Health Consensus Development Conference has
recommended that for stage I management, tumors that are stage IC, grade 3, or clear cell
require chemotherapy; they could not, however, reach a consensus for the other ovarian
cancer stage I subsets (33). These recommendations have not been revised since then.
Currently, all ovarian cancer histotypes are treated similarly, but it is becoming
increasingly clear that each should have different treatment approaches. As shown in
Table 2.5, there are not only differences in five-year survival by stage, but by histotype as
well, which could reflect histotype-specific features that impact overall prognoses and
treatment effects. Clear cell tumors have been consistently shown to have strong
22
resistance to platinum-based regimens with significantly poorer response rates (11-27%)
in comparison to serous tumors (73-81%) (34). In addition, advanced-stage clear cell
tumors tend to have low median survival. The mucinous histotype has also been shown to
be less responsive to platinum-based chemotherapy with worse progression-free and
overall survival (35). Randomized trials of patients with advanced ovarian cancer treated
with taxane/platinum regimens showed that the mucinous and clear cell histotypes were
independent predictors of poor prognosis (36, 37). Given these differential responses to
current treatments, novel, histotype-specific treatment approaches may be more
appropriate.

GENETIC AND ENVIRONMENTAL (“NON-GENETIC”) RISK FACTORS
Genetic factors
Ovarian cancer has a strong genetic basis. Women with a first-degree family
history of the disease are two to three times as likely to develop it, with increasing risk
with greater numbers of affected first- and second-degree relatives (38). Approximately
10% of all ovarian cancers can be associated with a genetic familial predisposition (2). A
large part is attributable to inherited mutations in high-penetrance susceptibility genes.
For example, there are two clinically distinct syndromes associated with hereditary
ovarian cancer. The more common of the two is the hereditary breast-ovarian cancer
syndrome, which is associated with germline mutations in BRCA1 and BRCA2 tumor
suppressor genes. The other is the hereditary non-polyposis colorectal cancer syndrome,
which is caused by germline mutations in mismatch repair genes (ie. MLH1 and MSH2
genes). Recent studies have also confirmed RAD51-related genes C and D as ovarian
23
cancer susceptibility genes. Six monoallelic pathogenic mutations in RAD51C have been
found to confer an increased risk for breast and ovarian cancer, and women who are
RAD51D mutation carriers are about six times as likely to develop ovarian cancer (39-
41). In addition, part of this heritability is due to several low-penetrance genetic variants.
Genome-wide association studies have identified 18 distinct ovarian cancer susceptibility
loci to date (42-48).
Given that 39% of women who inherit a BRCA1 mutation and approximately 11%
to 17% of women who inherit a BRCA2 mutation will develop ovarian cancer by the age
of 70 (49), many elect to have a risk-reducing prophylactic bilateral salpingo-
oophorectomy (RR-BSO) since its effectiveness in cancer prevention has been well-
documented. One multicenter study of women with disease-associated germline BRCA1
or BRCA2 mutations compared those who received a RR-BSO to those who did not and
found that the surgery reduced the risk of ovarian cancer by 96% (50). Women carrying
other major gene mutations that predispose to the disease, such as those involving DNA
mismatch repair genes or simply those with a high lifetime risk, may also elect to have a
RR-BSO.

Environmental ("non-genetic") factors
One proposed theory regarding the etiology of ovarian cancer is the incessant
ovulation hypothesis, which argues that repeated ovulation-induced trauma and repair of
the ovarian surface epithelium, without pregnancy-induced rest periods, contributes to
neoplasm formation (4). Hence, according to this theory, suppression of ovulation would
reduce a woman's risk of ovarian cancer, which could explain the protective effects of
24
oral contraceptive use, breastfeeding, and parity. In a large pooled analysis, the
Collaborative Group on Epidemiological Studies of Ovarian Cancer (Collaborative
Group) found a 43% reduction in risk with an average of 5.8 years of OC use, with
increased protection with longer duration of use (51). Increasing parity is also
consistently associated with a decreased risk of ovarian cancer (52). Pike et al showed
that women who had given birth once were 0.62 times as likely to develop the disease
relative to women who were nulliparous, and this protection became greater with each
additional birth (53). Breastfeeding has also been shown to reduce risk of ovarian cancer.
In the prospective Nurses’ Health Study, women who had breastfed at least 18 months
were 34% less likely to develop the disease in comparison to women who had not (54).
The incessant ovulation hypothesis assumes that ovarian cancer originates from
the ovarian surface epithelium and that the various subtypes arise due to metaplastic
changes that cause different cell types to develop. However, many now believe that
ovarian cancers are not ovarian in origin, but actually arise from different pelvic regions
as indicated in Table 2.4. This dramatic shift in thinking is likely due to the heterogeneity
of ovarian cancer as a disease, in which each subtype is not only characterized by distinct
molecular and genetic features, but epidemiologic risk factors as well. An analysis by
Pearce et al that jointly modeled several well-established ovarian cancer risk and
protective factors found that their effects on disease risk varied across the different
histotypes (52). This is clearly evident for endometriosis as studies have consistently
shown that the increased risk associated with endometriosis is predominantly seen among
the clear cell and endometrioid histotypes (55). Smoking has also been consistently found
to be positively associated with only risk of mucinous ovarian cancer (56). Table 2.6
25
presents the relationships between certain etiologic factors and each histotype based on
the results of large pooled analyses.
Another proposed hypothesis explaining the etiology of ovarian cancer is the role
of inflammation. The process of ovulation induces inflammatory reactions, which entail
cell damage, oxidative stress, and elevated levels of cytokines and prostaglandins, all
which could be potentially mutagenic (57). The minimizing of ovulation-induced
inflammation associated with OC use and parity could explain their protective effects. In
addition, the 30% decreased risk of ovarian cancer among those who had a tubal ligation
can also be explained by a diminishing of inflammation since the procedure would block
retrograde flow of inflammatory agents from the vagina to the peritoneal cavity (53, 58).
One pooled analysis found its effect to be significantly more protective for the
endometrioid and clear cell histotypes, which could be due to the location of the ligation,
which prevents the retrograde transport and ovarian seeding of cells from the
endometrium but not the distal tubes (59). In addition, endometriosis, which is thought to
be the precursor lesion for the clear cell and endometrioid histotypes, has been shown to
cause local pelvic inflammation. The implantation of ectopic endometrial tissue in
endometriosis is associated with inflammatory events, such as macrophage activation and
elevation of cytokines and growth factors (60). Given that the incessant ovulation
hypothesis does not appear to reconcile many of the epidemiologic associations we
observe for ovarian cancer, inflammation could be a contributing factor.
A more unifying hypothesis for these risk and protective factors is the effects of
the hormonal environment. The suppression of ovulation by previously mentioned
protective factors, such as OC use and parity, is also related to varying levels of estrogen
26
and progesterone. In addition, factors unrelated to ovulation, such as body mass index
(BMI) and height, have been found to be positively associated with ovarian cancer risk,
which could be attributable to the production of excess endogenous estrogens. A meta-
analysis conducted by the Collaborative Group found a 7% increased risk for every 5 cm
increase in height, and a separate analysis showed a 20% and 30% increased risk for
overweight (BMI=25-29.9 kg/m
2
) and adult obesity (BMI=30+ kg/m
2
), respectively,
relative to those classified as normal weight (BMI=18.5-24.9 kg/m
2
) (61, 62). Obesity
has also been found to be a well-established risk factor for other hormone-related cancers,
such as postmenopausal breast cancer, which suggests the carcinogenic impact of
hormone elevations (63). Hormone therapy is another potential risk factor for ovarian
cancer that is directly related to hormone levels, and it will be the primary focus of this
dissertation.

27
Table 2.1. Three Main Staging Systems for Ovarian Cancer

FIGO
Stage
AJCC/TNM SEER Description
IA T1a, N0, M0 Localized One ovary, capsule intact
IB T1b, N0, M0 Localized Both ovaries, capsule intact
IC T1c, N0, M0 Regional
Capsule ruptured, cancer on surface,
malignant cells in ascites or peritoneal
washings
IIA T2a, N0, M0 Regional Uterus, fallopian tubes
IIB T2b, N0, M0 Regional Bladder, sigmoid colon, rectum
IIIA1 T1/T2, N1, M0 Distant Retroperitoneal lymph nodes
IIIA2
T3a2, N0/N1,
M0
Distant
Tiny deposits of cancer in lining of upper
abdomen
IIIB T3b, N0/N1, M0 Distant
Larger deposits of cancer ≤2cm in the
abdomen
IIIC T3c, N0/N1, M0 Distant
Larger deposits of cancer >2cm in the
abdomen
IVA
any T, any N,
M1
Distant Fluid around the lungs
IVB
any T, any N,
M1
Distant
Inside of spleen or liver and/or to other
tissues/organs outside the peritoneal cavity
Abbreviations: FIGO, International Federation of Gynecology and Obstetrics;
AJCC/TNM, American Joint Committee on Cancer/Tumor, Node, Metastasis; SEER,
Surveillance, Epidemiology, and End Results.

28
Table 2.2. Overall FIGO Stage Distribution and 5-year Relative Survival by Stage at
Diagnosis for Ovarian Cancer, 2000-2012, All Races Combined

Overall
FIGO Stage
Description
Stage
Distribution (%)
a

5-year Relative
Survival (%)
a

I Growth limited to the ovaries 25.4% 85.5%
II
Growth involving one or both
ovaries; with pelvic extension
8.9% 65.4%
III
Tumor involving one or both
ovaries with peritoneal
implants outside the pelvis
and/or positive
retroperitoneal or inguinal
nodes
37.4% 36.8%
IV
Growth involving one or both
ovaries, with distant
metastases
28.3% 16.4%
Abbreviations: FIGO, International Federation of Gynecology and Obstetrics.
a
Calculated using SEER*Stat Software Version 8.2.1 (8).

29
Table 2.3. Common Grading Systems for Ovarian Cancer

FIGO MDACC
None
Specified
a

Description
Low Low 1
Well-differentiated cells, slower-
growing, less likely to spread
Intermediate High 2
Moderately differentiated cells that
look more abnormal, slightly faster-
growing
High High 3
Poorly differentiated cells that are
abnormal-looking, fast-growing
High High 4
Undifferentiated cells that are
abnormal-looking, fast-growing,
very likely to spread
Abbreviations: FIGO, International Federation of Gynecology and Obstetrics; MDACC,
M.D. Anderson Cancer Center.
a
Tumors are typically graded 1 to 4 depending on the amount of abnormality if a grading
system is not specified (12).

30
Table 2.4. Distribution and Characteristics of the Main Ovarian Cancer Histotypes

Histotype
Percent
Diagnosed
(13)
a

Hypothesized
Origin (15)
Molecular Genetic
Alterations (64)
High-grade
Serous
70%
Fimbriated portion of
the fallopian tube
TP53 mutations,
Overexpression of
HER2/neu and AKT2,
Inactivation of p16 gene
Low-grade
Serous
2%
Fimbriated portion of
the fallopian tube
BRAF mutations,
KRAS mutations
Mucinous 2-4%
Transitional-type
epithelium at tubal-
mesothelial junction
KRAS mutations
Endometrioid 10% Endometriosis
Loss of heterozygosity
or PTEN mutations,
β-catenin gene mutations,
KRAS mutations,
Microsatellite instability
Clear Cell 10% Endometriosis
KRAS mutations,
Microsatellite instability,
TGF-β RII mutations
a
Does not total 100% due to other tumor types that are not classified as one of the five
main histotypes.

31
Table 2.5. Five-year Ovarian Cancer Survival Rates by Histotype and FIGO Stage at
Diagnosis

FIGO
Stage
a

High-grade
Serous
b

Low-grade
Serous
b

Mucinous
b
Endometrioid
b

Clear
Cell
b

I 79.9% 92.1% 86.5% 89.9% 83.8%
IA 82.6% 88.9% 88.4% 90.6% 84.7%
IB 87.2%
c
77.2% 92.1% 86.3%
IC 76.8% 97.2% 82.1% 89.0% 81.5%
II 65.0% 64.3% 66.4% 80.3% 66.3%
IIA
d
67.1%
c
60.9% 83.7% 67.2%
IIB 66.3% 50.0% 60.8% 81.2% 72.5%
III 38.4% 60.0% 26.2% 54.4% 31.0%
IIIA 50.9% 100.0% 43.5% 68.6% 45.0%
IIIB 50.3% 35.6% 11.6% 64.9% 25.7%
IIIC 37.7% 59.9% 31.8% 53.0% 29.2%
IV
d
22.2% 27.2% 8.6% 29.0% 12.7%
Abbreviations: FIGO, International Federation of Gynecology and Obstetrics.
a
Five-year survival rates for overall FIGO stages I through IV include tumors that were
not specified as A, B, or C.
b
Calculated using SEER*Stat Software Version 8.2.1 (8).
c
Could not be calculated.
d
FIGO stages IIA1 and IIA2 have been combined. FIGO stages IVA and IVB have been
combined.

32
Table 2.6. Type of Association Between Each Etiologic Factor and Ovarian Cancer Risk
by Histotype Based on Large Pooled Analyses

Etiologic
Factor
High-grade
Serous
a

Low-grade
Serous
a

Mucinous
a

Endom-
etrioid
a

Clear
Cell
a

OC Use (52) - - - - -
Parity (52) - - - - -
Tubal Ligation
(59)
- - - - -
Endometriosis
(55)
b

None + None + +
Family History
(52)
+ + + + +
Smoking (56) None None + None None
Obesity (65)
b
None + + + None
Abbreviations: OC, oral contraceptive.
a
“+” refers to a positive association; “-” refers to a negative association.
b
Analyses for serous tumors were differentiated by grade. The associations presented for
the other risk factors were the results for all serous tumors combined.

33
Figure 2.1. Age-adjusted Incidence and Mortality Rates
a
for Ovarian Cancer, 2000 to
2012, All Races Combined

a
Derived from SEER*Stat Software Version 8.2.1 (8).

0
2
4
6
8
10
12
14
16
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Age-adjusted Rates (per 100,000)
Year
Incidence
Mortality
14.3
8.9
11.9
5
7.4
34
REFERENCES

1. American Cancer Society. Ovarian Cancer.
(http://www.cancer.org/acs/groups/cid/documents/webcontent/003130-pdf.pdf).
(Accessed February 2, 2016).
2. Jhamb N, Lambrou NC. Epidemiology and Clinical Presentation of Ovarian
Cancer. In: Bristow RE, Armstrong DK, eds. Ovarian Cancer. 1 ed. Philadephila,
PA: Saunders Elsevier; 2010.
3. National Cancer Institute. "SEER Cancer Statistics Review, 1975-2012."
(http://seer.cancer.gov/csr/1975_2012_pops09/). (Accessed February 2, 2016).
4. Ho SM. Estrogen, progesterone and epithelial ovarian cancer. Reprod Biol
Endocrinol. 2003;1:73.
5. Ferriss JS, King ER, Modesitt SC. Borderline Epithelial Ovarian Tumors, Sex
Cord-Stromal Tumors, and Germ Cell Tumors. In: Bristow RE, Armstrong DK,
eds. Ovarian Cancer. 1 ed. Philadelphia, PA: Saunders Elsevier; 2010.
6. Karam AK, Walsh C, Karlan BY. Ovarian Cancer Surgery. In: Bristow RE,
Armstrong DK, eds. Ovarian Cancer. 1 ed. Philadelphia, PA: Saunders Elsevier;
2010.
7. Le T, Adolph A, Krepart GV, et al. The benefits of comprehensive surgical
staging in the management of early-stage epithelial ovarian carcinoma. Gynecol
Oncol. 2002;85(2):351-355.

35
8. Surveillance, Epidemiology, and End Results (SEER) Program Research Data
(1973-2012) Version 8.2.1. National Cancer Institute, DCCPS, Surveillance
Research Program, Surveillance Systems Branch; 2012.
(http://seer.cancer.gov/seerstat/software/) (Accessed February 2, 2016).
9. Matsuno RK, Sherman ME, Visvanathan K, et al. Agreement for tumor grade of
ovarian carcinoma: analysis of archival tissues from the surveillance,
epidemiology, and end results residual tissue repository. Cancer Causes Control.
2013;24(4):749-757.
10. Malpica A, Deavers MT, Lu K, et al. Grading ovarian serous carcinoma using a
two-tier system. Am J Surg Pathol. 2004;28(4):496-504.
11. Kurman RJ, Visvanathan K, Roden R, et al. Early detection and treatment of
ovarian cancer: shifting from early stage to minimal volume of disease based on a
new model of carcinogenesis. Am J Obstet Gynecol. 2008;198(4):351-356.
12. Cancer AJCo. AJCC Cancer Staging Manual. 7th ed. New York, NY: Springer;
2010.
13. Gilks CB. Molecular abnormalities in ovarian cancer subtypes other than high-
grade serous carcinoma. J Oncol. 2010;2010:740968.
14. Kurian AW, Balise RR, McGuire V, et al. Histologic types of epithelial ovarian
cancer: have they different risk factors? Gynecol Oncol. 2005;96(2):520-530.
15. Kurman RJ, Shih Ie M. The origin and pathogenesis of epithelial ovarian cancer:
a proposed unifying theory. Am J Surg Pathol. 2010;34(3):433-443.
16. Hennekens CH, Buring JE. Epidemiology in Medicine. 1 ed: Little, Brown and
Co.; 1987.
36
17. Brown PO, Palmer C. The preclinical natural history of serous ovarian cancer:
defining the target for early detection. PLoS Med. 2009;6(7):e1000114.
18. Rosenthal AN, Fraser L, Manchanda R, et al. Results of annual screening in phase
I of the United Kingdom familial ovarian cancer screening study highlight the
need for strict adherence to screening schedule. J Clin Oncol. 2013;31(1):49-57.
19. Buys SS, Partridge E, Black A, et al. Effect of screening on ovarian cancer
mortality: the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening
Randomized Controlled Trial. JAMA. 2011;305(22):2295-2303.
20. Moyer VA, U.S. Preventive Services Task Force. Screening for ovarian cancer:
U.S. Preventive Services Task Force reaffirmation recommendation statement.
Ann Intern Med. 2012;157(12):900-904.
21. Goff BA, Lowe KA, Kane JC, et al. Symptom triggered screening for ovarian
cancer: a pilot study of feasibility and acceptability. Gynecol Oncol.
2012;124(2):230-235.
22. Rufford BD, Jacobs IJ, Menon U. Feasibility of screening for ovarian cancer
using symptoms as selection criteria. BJOG. 2007;114(1):59-64.
23. Twombly R. Cancer killer may be "silent" no more. J Natl Cancer Inst.
2007;99(18):1359-1361.
24. Goff BA, Mandel LS, Drescher CW, et al. Development of an ovarian cancer
symptom index: possibilities for earlier detection. Cancer. 2007;109(2):221-227.
25. Rossing MA, Wicklund KG, Cushing-Haugen KL, et al. Predictive value of
symptoms for early detection of ovarian cancer. J Natl Cancer Inst.
2010;102(4):222-229.
37
26. Goff BA. Advanced ovarian cancer: what should be the standard of care? J
Gynecol Oncol. 2013;24(1):83-91.
27. Elattar A, Bryant A, Winter-Roach BA, et al. Optimal primary surgical treatment
for advanced epithelial ovarian cancer. Cochrane Database Syst Rev.
2011(8):CD007565.
28. Bell JG, Lutman CV. Chemotherapy for Ovarian Cancer. In: Bristow RE,
Armstrong DK, eds. Ovarian Cancer. 1 ed. Philadelphia, PA: Saunders Elsevier;
2010.
29. Ushijima K. Treatment for recurrent ovarian cancer-at first relapse. J Oncol.
2010;2010:497429.
30. Vergote I, Trope CG, Amant F, et al. Neoadjuvant chemotherapy or primary
surgery in stage IIIC or IV ovarian cancer. N Engl J Med. 2010;363(10):943-953.
31. Kang S, Nam BH. Does neoadjuvant chemotherapy increase optimal
cytoreduction rate in advanced ovarian cancer? Meta-analysis of 21 studies. Ann
Surg Oncol. 2009;16(8):2315-2320.
32. Trimbos JB, Vergote I, Bolis G, et al. Impact of adjuvant chemotherapy and
surgical staging in early-stage ovarian carcinoma: European Organisation for
Research and Treatment of Cancer-Adjuvant ChemoTherapy in Ovarian
Neoplasm trial. J Natl Cancer Inst. 2003;95(2):113-125.
33. National Institutes of Health Consensus Development Conference Statement.
Ovarian cancer: screening, treatment, and follow-up. Gynecol Oncol. 1994;55(3
Pt 2):S4-14.
38
34. Pectasides D, Pectasides E, Psyrri A, et al. Treatment issues in clear cell
carcinoma of the ovary: a different entity? Oncologist. 2006;11(10):1089-1094.
35. Hess V, A'Hern R, Nasiri N, et al. Mucinous epithelial ovarian cancer: a separate
entity requiring specific treatment. J Clin Oncol. 2004;22(6):1040-1044.
36. Mackay HJ, Brady MF, Oza AM, et al. Prognostic relevance of uncommon
ovarian histology in women with stage III/IV epithelial ovarian cancer. Int J
Gynecol Cancer. 2010;20(6):945-952.
37. Winter WE, 3rd, Maxwell GL, Tian C, et al. Prognostic factors for stage III
epithelial ovarian cancer: a Gynecologic Oncology Group Study. J Clin Oncol.
2007;25(24):3621-3627.
38. Auranen A, Pukkala E, Makinen J, et al. Cancer incidence in the first-degree
relatives of ovarian cancer patients. Br J Cancer. 1996;74(2):280-284.
39. Meindl A, Hellebrand H, Wiek C, et al. Germline mutations in breast and ovarian
cancer pedigrees establish RAD51C as a human cancer susceptibility gene. Nat
Genet. 2010;42(5):410-414.
40. Loveday C, Turnbull C, Ramsay E, et al. Germline mutations in RAD51D confer
susceptibility to ovarian cancer. Nat Genet. 2011;43(9):879-882.
41. Wickramanyake A, Bernier G, Pennil C, et al. Loss of function germline
mutations in RAD51D in women with ovarian carcinoma. Gynecol Oncol.
2012;127(3):552-555.
42. Bolton KL, Tyrer J, Song H, et al. Common variants at 19p13 are associated with
susceptibility to ovarian cancer. Nat Genet. 2010;42(10):880-884.
39
43. Goode EL, Chenevix-Trench G, Song H, et al. A genome-wide association study
identifies susceptibility loci for ovarian cancer at 2q31 and 8q24. Nat Genet.
2010;42(10):874-879.
44. Song H, Ramus SJ, Tyrer J, et al. A genome-wide association study identifies a
new ovarian cancer susceptibility locus on 9p22.2. Nat Genet. 2009;41(9):996-
1000.
45. Shen H, Fridley BL, Song H, et al. Epigenetic analysis leads to identification of
HNF1B as a subtype-specific susceptibility gene for ovarian cancer. Nat
Commun. 2013;4:1628.
46. Pharoah PD, Tsai YY, Ramus SJ, et al. GWAS meta-analysis and replication
identifies three new susceptibility loci for ovarian cancer. Nat Genet.
2013;45(4):362-370.
47. Bojesen SE, Pooley KA, Johnatty SE, et al. Multiple independent variants at the
TERT locus are associated with telomere length and risks of breast and ovarian
cancer. Nat Genet. 2013;45(4):371-384.
48. Permuth-Wey J, Lawrenson K, Shen HC, et al. Identification and molecular
characterization of a new ovarian cancer susceptibility locus at 17q21.31. Nat
Commun. 2013;4:1627.
49. Chen S, Parmigiani G. Meta-analysis of BRCA1 and BRCA2 penetrance. J Clin
Oncol. 2007;25(11):1329-1333.
50. Rebbeck TR, Lynch HT, Neuhausen SL, et al. Prophylactic oophorectomy in
carriers of BRCA1 or BRCA2 mutations. N Engl J Med. 2002;346(21):1616-
1622.
40
51. Collaborative Group on Epidemiological Studies of Ovarian C, Beral V, Doll R,
et al. Ovarian cancer and oral contraceptives: collaborative reanalysis of data from
45 epidemiological studies including 23,257 women with ovarian cancer and
87,303 controls. Lancet. 2008;371(9609):303-314.
52. Pearce CL, Rossing MA, Lee AW, et al. Combined and interactive effects of
environmental and GWAS-identified risk factors in ovarian cancer. Cancer
Epidemiol Biomarkers Prev. 2013;22(5):880-890.
53. Pike MC, Pearce CL, Peters R, et al. Hormonal factors and the risk of invasive
ovarian cancer: a population-based case-control study. Fertil Steril.
2004;82(1):186-195.
54. Danforth KN, Tworoger SS, Hecht JL, et al. Breastfeeding and risk of ovarian
cancer in two prospective cohorts. Cancer Causes Control. 2007;18(5):517-523.
55. Pearce CL, Templeman C, Rossing MA, et al. Association between endometriosis
and risk of histological subtypes of ovarian cancer: a pooled analysis of case-
control studies. Lancet Oncol. 2012;13(4):385-394.
56. Faber MT, Kjaer SK, Dehlendorff C, et al. Cigarette smoking and risk of ovarian
cancer: a pooled analysis of 21 case-control studies. Cancer Causes Control.
2013;24(5):989-1004.
57. Ness RB, Cottreau C. Possible role of ovarian epithelial inflammation in ovarian
cancer. J Natl Cancer Inst. 1999;91(17):1459-1467.
58. Cibula D, Widschwendter M, Majek O, et al. Tubal ligation and the risk of
ovarian cancer: review and meta-analysis. Hum Reprod Update. 2011;17(1):55-
67.
41
59. Sieh W, Salvador S, McGuire V, et al. Tubal ligation and risk of ovarian cancer
subtypes: a pooled analysis of case-control studies. Int J Epidemiol.
2013;42(2):579-589.
60. Ness RB. Endometriosis and ovarian cancer: thoughts on shared pathophysiology.
Am J Obstet Gynecol. 2003;189(1):280-294.
61. Olsen CM, Green AC, Whiteman DC, et al. Obesity and the risk of epithelial
ovarian cancer: a systematic review and meta-analysis. Eur J Cancer.
2007;43(4):690-709.
62. Collaborative Group on Epidemiological Studies of Ovarian C. Ovarian cancer
and body size: individual participant meta-analysis including 25,157 women with
ovarian cancer from 47 epidemiological studies. PLoS Med. 2012;9(4):e1001200.
63. Cleary MP, Grossmann ME. Minireview: Obesity and breast cancer: the estrogen
connection. Endocrinology. 2009;150(6):2537-2542.
64. Jinawath N, Shih I-M. Biology and Pathology of Ovarian Cancer. In: Bristow RE,
Armstrong DK, eds. Ovarian Cancer. 1 ed. Philadelphia, PA: Saunders Elsevier;
2010.
65. Olsen CM, Nagle CM, Whiteman DC, et al. Obesity and risk of ovarian cancer
subtypes: evidence from the Ovarian Cancer Association Consortium. Endocr
Relat Cancer. 2013;20(2):251-262.

42
Chapter 3: Background on Hormone Therapy

43
Abbreviation List:
BMI – body mass index
cEPT – continuous estrogen-progestin combined therapy
CI – confidence interval
EE – conjugated equine estrogens
EPT – estrogen-progestin combined therapy
ET – estrogen-only therapy
HERS – Heart and Estrogen/Progestin Replacement Study
HT – hormone therapy
MPA – medroxyprogesterone acetate
OC – oral contraceptive
RR – relative risk
sEPT – sequential estrogen-progestin combined therapy
U.S. – United States
WHI – Women’s Health Initiative

Running Head:
Background on Hormone Therapy
44
Hormone therapy (HT) supplements a lack of naturally occurring hormones
among menopausal women for climacteric symptom relief and osteoporosis prevention.
According to the Agency for Healthcare Research Quality, there were close to 18 million
HT users before the Women's Health Initiative (WHI) randomized trial in 2002 (1), and
while this number has declined, there are still a significant number of women using HT
today. Given that more women are expected to spend at least a third of their life
postmenopausally as life expectancy increases, it is important that we understand how the
most effective treatment for menopausal symptoms may impact risk of the most fatal
female gynecologic cancer.

OVERVIEW OF HORMONE THERAPY
Historical trends and statistics
The popularity of HT in the form of estrogen-only therapy (ET) reached its peak
in 1975 (2), but significantly declined when studies implicating a causal relationship
between ET use and endometrial cancer were first published (3). A few years later, the
addition of a progestin in the form of estrogen-progestin combined therapy (EPT) was
found to counteract the proliferative effect of exogenous estrogen on the postmenopausal
endometrium (4), prompting a steady increase in HT use again as EPT offered
menopausal symptom relief without risk of endometrial cancer (2).
This increase in HT use was also attributable to several observational studies
suggesting an effect on lipid profiles that decreased risk of coronary heart disease (5, 6).
Studies had shown that not only did premenopausal women have a lower risk and
incidence of hypertension and cardiovascular disease compared to age-matched men, but
45
that this benefit diminished after menopause, which led to the notion that sex hormones,
such as estrogen, played some cardioprotective role that HT use could maintain (7, 8).
One cohort study reported a 39% decreased risk of major coronary events among current
HT users relative to non-users (9). In addition, several clinical trials noted favorable lipid
profile changes among HT users, with reductions in low-density lipoprotein levels and
increases in high-density lipoprotein levels (10).
However, the early termination of the WHI quickly reversed this trend as the
number of HT prescriptions dropped by 38% just one year after (11). The WHI found
higher rates of thromboembolic events, breast cancer, and coronary heart disease,
contrary to the results of previous studies, concluding that overall health risks associated
with EPT use exceeded its benefits (12). The Heart and Estrogen/Progestin Replacement
Study (HERS I) and its follow-up, HERS II, also found similar results, suggesting that
HT use may actually increase risk of cardiovascular disease in postmenopausal women
(13, 14). However, one explanation for these conflicting results could be the timing of
HT use relative to menopause. The "timing hypothesis" proposed that early initiation of
HT (ie. during the perimenopausal transition) may inhibit atherosclerosis, but that these
beneficial effects may be lost if use began several years after menopause when the
atherosclerosis already progressed to a complicated plaque stage (15). Another
explanation could simply be the starting age of HT use. In the California Teachers Study,
Stram et al found reduced risks of ischemic heart disease mortality among younger HT
users, but not among older postmenopausal women, even if HT use began close to
menopause (16). Hence, the original observational studies that suggested a
cardioprotective effect may be valid. However, the results of the WHI, HERS I, and
46
HERS II had a lasting impact as HT use declined from 17.9 million users in 2001 to 5.8
million users in 2008 as shown in Figure 3.1 (1).

Hormone therapy options
As previously mentioned, there are two main types of HT: ET, which includes
only an estrogen component, and EPT, which includes a progestin component in addition
to the estrogen. Use of unopposed estrogen has been shown to promote hyperplasia of
cells in the uterine lining, hence women with intact uteri needing menopausal symptom
relief are strongly recommended to take EPT to minimize risk of endometrial cancer (17,
18). The brands Premarin, which consists of conjugated equine estrogens (CEE), and
Provera, which consists of medroxyprogesterone acetate (MPA), are the most frequently
used synthetic estrogen and progestin, respectively, in the United States (U.S.) (19). The
combination of CEE at 0.625 mg/day and MPA at 2.5 mg/day has been considered the
minimum effective dose for bone preservation and endometrium protection and is one of
the most frequently prescribed EPT preparations (20); however, thoughts on this are
changing due to growing interest in lower-dosed HT to minimize risk of breast cancer
and other hormone-related adverse effects.
Because of the clear dose-response relationship between exogenous estrogen use
and risk of adverse endometrial effects, several trials have looked at the progestin dose
needed for adequate endometrial protection when lower estrogen doses are used, with
most demonstrating an excellent endometrial safety profile even with concurrent use of
lower doses of progestin (21). In addition, these lower hormone doses were shown to still
be effective in relieving menopausal symptoms. The Women's Health, Osteoporosis,
47
Progestin, Estrogen study found that CEE at 0.45 and 0.3 mg/day and MPA at 1.5 mg/day
still decreased the number and severity of hot flushes and improved vaginal atrophy
during a one-year period (22). Studies have also shown their promise in minimizing
postmenopausal bone loss (21). However, little is known about their long-term effects,
especially with regard to HT-associated conditions such as breast cancer.
There are also two main EPT regimens based on the frequency of progestin use.
Sequential estrogen-progestin combined therapy (sEPT) consists of taking the estrogen
component every day while adding the progestin 5 to 15 days of a woman's menstrual
cycle whereas continuous estrogen-progestin combined therapy (cEPT) consists of taking
both components daily (23). Because cEPT includes daily progestin use, it typically has a
lower progestin dose than sEPT. In the U.S., the most common sEPT includes 5 to 10
mg/day of MPA for 10 days per 28-day cycle whereas for cEPT, the dose is 2.5 mg/day
of MPA daily (24). However, in some European countries such as Sweden, a similar
daily dose of progestin is used for both EPT regimens (24). These variations in dosing
and scheduling, which are shown in Table 3.1, may result in different cumulative
amounts of progestin. This could explain the inconsistent results seen across various
ovarian cancer studies, which are described below in the “EPT and ovarian cancer risk”
section.
The decision to use sEPT versus cEPT usually depends on a woman’s personal
preference, menstruation history, and menopausal status. sEPT mimics a more natural
hormone cycle, which induces withdrawal bleeding when progestin use is ceased, causing
many of the side effects associated with menstruation. Hence, cEPT has been
recommended as a way to reduce or eliminate this bleeding. However, there are concerns
48
regarding the adverse impact of continuous progestin use. While ET use has been shown
to result in favorable lipid level effects as previously mentioned, some studies found
these effects to diminish with progestin use (25, 26). In addition, almost all published
studies to date have found greater breast cancer risk with EPT use than ET use (27). One
meta-analysis showed that while ET use was associated with a 16% increased risk of
breast cancer, the increased risk associated with EPT use was 39% (28).
Women who choose not to use conventional, synthetic HT may instead take
bioidentical hormones or other alternative medicines for menopausal symptom relief.
Bioidentical hormones are molecularly identical to the hormones produced endogenously,
and for HT, they include micronized estrone, estradiol, estriol, and progesterone, natural
forms of the hormones typically derived from plants. They are also commonly
compounded by pharmacists to produce preparations individualized to the patient.
Herbals or nutritional supplements consisting of phytoestrogens are commonly used as
well. These HT substitutes have been found to be effective for menopausal symptom
relief, but their long-term effects are unknown due to limited scientific data. Plant foods
are thought to be safe, but the isolated and often concentrated components have yet to be
evaluated for long-term use (29). This dissertation will focus on traditional, synthetic HT
since it is more common and has more adequate product quality control.

Role of hormone therapy in ovarian carcinogenesis
Because hormones play a central role in regulating cell proliferation,
differentiation, and apoptosis, they have been thought to influence ovarian cancer
development since dysregulation of any of these processes could allow cells with
49
harbored mutations in proto-oncogenes and tumor suppressor genes to thrive (30). In
addition, the well-established protective effects of oral contraceptive (OC) use and parity
may be beyond just ovulation suppression in which case additional hormonal
mechanisms may play an etiologic role.

Estrogen. Estrogens have long been implicated as etiologic factors of ovarian
cancer. Their levels in the ovarian tissue have been shown to be at least 100 times higher
than circulating levels with levels in the follicular fluid of ovulatory follicles being even
greater, suggesting the role of a high estrogen climate in ovarian carcinogenesis (31). In
relation to the long-standing incessant ovulation hypothesis, it was noted that cells of the
ovarian surface epithelium undergo rapid proliferation during the 24 hours after ovulation
(32). In fact, this epithelium proliferates at times when estrogenic influences are greater
during the menstrual cycle, and this increased activity due to the estrogen-rich
environment has been thought to enhance the risk of mutations. In vitro studies have
shown estrogen’s ability to stimulate the proliferation of ovarian surface epithelium cells
(33), and animal studies involving the treatment of guinea pigs with estradiol have also
resulted in the development of serous ovarian cysts (34).
As discussed in Chapter 2, there has been growing evidence against the ovarian
surface epithelium as the site of origin for ovarian carcinogenesis (35). In fact, the
heterogeneity of ovarian cancer includes different cells of origin for its main histotypes,
but the role of estrogen in these various cells has been noted as well. The fallopian tube
fimbriae, which have recently been hypothesized to be the origin of serous tumors, have
been shown to exhibit proliferative activity in the epithelium during the menstrual cycle,
50
and this is thought to be induced by estrogen (36). In addition, endometrioid and clear
cell tumors are thought to arise from endometriosis, an estrogen-dependent gynecologic
condition consisting of endometriotic lesions that have been shown to harbor molecular
aberrations that favor increased local production of estradiol (30, 35). There also appears
to be a plateau in the incidence of endometrioid, clear cell, and mucinous tumors after
menopause, which suggests that ovarian estrogens may play a role in the development of
these tumors; interestingly, the incidence of serous tumors continues to increase well
after menopause (30).
The risk-inducing effects of estrogen have also been seen in epidemiologic
associations related to increased estrogen levels. Body mass index (BMI) has been shown
to be positively associated with ovarian cancer risk, likely as a result of excess
endogenous estrogens synthesized in the adipose tissue through androgen aromatization
(37, 38). This would suggest that the association between obesity and ovarian cancer risk
is weaker among HT users since they would already exhibit high circulating estrogen
levels in comparison to non-HT users, and this has been observed (39, 40). While the data
demonstrating this effect modification were sparse in the cohort study by Leitzmann et al
(39), the risk estimates for the association between BMI and ovarian cancer risk were
consistently higher among those who had never used HT. Women who had a BMI of 30+
kg/m
2
and had never used HT were 1.83 times as likely to develop ovarian cancer;
however, if they had used HT, there was no increased risk at all (39).

Progesterone. Progesterone, on the other hand, has been thought to protect
against ovarian tumor development. Its levels during pregnancy, a well-established
51
protective factor for ovarian cancer, are significantly elevated (32). For example, the
increase in human chorionic gonadotropin during the first few weeks of pregnancy
stimulates the corpus luteum to continue producing progesterone and not regress.
Progesterone production is continued with the synthesis of the placenta, which has been
shown to cause a 10-fold increase in maternal circulating progesterone levels (41). These
eight to nine months of elevated levels of progesterone may be an additional aspect that
contributes to the protection afforded by pregnancy, beyond simply ovulation
suppression. The greater protective effect of multiple gestation seen by some studies adds
further evidence as progesterone levels during multiple pregnancies have been shown to
be higher than during singleton pregnancies (42).
While endogenous progesterone synthesis decreases in OC users, the protection
afforded by OC use could be a result of the strong potency of synthetic progestins.
Synthetic 19-nortestosterone progestins are at least 100 times more potent than that of
progesterone, which could result in a high net progestational environment (43). In
addition, progestin-only OC formulations have been shown to incompletely suppress
ovulation, with up to 40% of users maintaining regular ovarian function (44). However,
use of such OCs was found to be associated with a significant decreased risk of ovarian
cancer comparable to the reduced risks associated with traditional formulations (44).
Given their similar degrees of protection, ovarian suppression does not appear to fully
explain the protective effect of OCs and suggests a potential role of progestin.
In addition, in vitro studies have shown progesterone’s inhibitory action on
ovarian cancer cell growth, with some even showing a clear induction of apoptosis of
these cells (31). This could be due to alterations in apoptosis regulators, which are seen
52
with high doses of progesterone. Animal studies have also shown that macaques treated
with progestins exhibited increased apoptosis in the ovarian cells (45), and laying hens
injected with MPA had a reduced frequency of spontaneous development of reproductive
tract adenocarcinoma, including ovarian cancer (46). It is also important to note that the
effects of progesterone may be related to the expression levels of their receptors. Using
semiquantitative, reverse transcription polymerase chain reaction, Lau et al showed a
significant downregulation of PR expression in ovarian cancer cells (47). Similarly,
another study from the Ovarian Tumor Tissue Analysis consortium found that high-grade
serous carcinomas whose tumors stained strongly positive for PR had improved survival
compared to those that stained negative or weakly positive (48). Hence, these results
collectively suggest an anti-carcinogenic effect of progesterone that is likely beyond the
inhibition of ovulation.

CRITICAL EVALUATION OF THE HORMONE THERAPY-OVARIAN CANCER
LITERATURE
Given the importance of HT on women's health and the role of hormones in
ovarian carcinogenesis, a better characterization of the HT-ovarian cancer association is
needed. Women have been using HT for decades, but, as previously mentioned, its use
has significantly declined due to the increased risk of breast cancer found in the WHI
(12). Since ovarian cancer shares similar etiologic factors with breast cancer, HT has
been suggested to be a potential ovarian cancer risk factor as well.

53
Estrogen-only therapy and ovarian cancer risk
Most epidemiologic studies evaluating this relationship have shown that use of
ET is associated with an increased risk of ovarian cancer. A recent pooled analysis of 52
studies by the Collaborative Group on Epidemiological Studies of Ovarian Cancer
(Collaborative Group) found a 32% increased risk with current ET use (relative risk
(RR)=1.32, 95% confidence interval (CI): 1.23, 1.41) (49). Similar results were seen in a
prior comprehensive meta-analysis of 14 studies through 2007 by Pearce et al, the
majority of which overlapped with the Collaborative Group’s analysis. They found every
five years of ET use to be associated with a 22% increased risk of ovarian cancer
(RR=1.22, 95% CI: 1.18, 1.27); Figure 3.2 shows the remarkable consistency of the
effect estimates across the studies included (50).
However, an important unanswered question with respect to ET use and ovarian
cancer is whether this association exists across all disease histotypes. While this has been
evaluated in seven studies (49, 51-56), several of them suffered from small sample sizes,
which has resulted in some inconsistency in the findings (see Tables 3.2 and 3.3). For
example, while some studies, including the Collaborative Group’s pooled analysis, have
found ET use to be associated with significant elevated risks for both serous and
endometrioid ovarian cancer (49, 52, 54), others have seen an increased risk for serous
tumors only (56). Interestingly, an analysis that looked at temporal trends in ovarian
cancer incidence and post-WHI HT use showed statistically significant period changes
around 2002 for both serous and endometrioid histotypes; among women who were 50
years or older, the age-standardized ovarian cancer incidence declined by 0.8% per year
54
before the WHI, but after the report, the rate declined by 2.4% per year, with the largest
changes found for endometrioid tumors (57).
Other characteristics of the ET-ovarian cancer association that should be
evaluated include ET’s duration and recency of use effects. Most individual studies have
shown greater risk associated with longer durations of ET use, with most of the increased
risk restricted to current users if recency was considered (58-60). However, the
Collaborative Group’s analysis, which is likely to be the most comprehensive analysis to
date, found no difference between short-term and long-term current HT use; however,
they did not report on duration according to HT type (49). Therefore, a complete
evaluation of the ET-ovarian cancer association that considers all of these features is
presented in Chapter 4 of this dissertation.

Estrogen-progestin combined therapy and ovarian cancer risk
Unlike ET, the association between EPT use and risk of ovarian cancer is less
clear. The Collaborative Group observed a 25% increased risk of ovarian cancer with
current EPT use (RR=1.25, 95% CI: 1.16, 1.34) (49), an effect estimate that appears to be
of greater magnitude than the one observed in Pearce et al’s meta-analysis, which showed
only a 10% increased risk for every five years of EPT use (RR=1.10, 95% CI 1.04, 1.16)
(50). In addition, the study-specific effect estimates in Pearce et al’s meta-analysis are
relatively inconsistent as can be seen in Figure 3.3; the Collaborative Group did not
present their results by study. There were four individual HT studies published after
Pearce et al’s meta-analysis, but included in the Collaborative Group’s report (58, 61-63).
All four showed an increased risk of ovarian cancer with EPT use, albeit of very different
55
magnitudes ranging from 1.08 to 1.50 with only two of the studies reaching statistical
significance. However, other previously published studies have not seen an increased risk
(59), which may not be surprising given the hypothesized biological role of progesterone.
Part of this discrepancy may be attributable to specific features of EPT use,
including its recency and duration. Few studies have considered both aspects
simultaneously, but those that did either found an increased risk among mainly current,
long-term EPT users or no risk at all (58, 59); these inconsistent findings are likely due to
limited numbers. In addition, as discussed previously, there are various types of EPT
regimens since the progestin component can be delivered either sequentially (sEPT) or
continuously (cEPT). The effects of sEPT and cEPT regimens on ovarian cancer risk has
not been extensively examined due to most studies having collected limited HT data.
Table 3.4 presents the current literature on this research question based on my review of
published studies. A meta-analysis of these results assessing use of each EPT regimen
dichotomously shows effect estimates (95% CIs) of 1.33 (1.22, 1.45) for sEPT and 1.18
(1.06, 1.31) for cEPT with a P-value for heterogeneity of 0.08. However, there was
significant heterogeneity across the six studies for both EPT regimens (P-value for
heterogeneity=0.02 for sEPT and P-value for heterogeneity=0.01 for cEPT).
Although a statistically significant difference between the two EPT regimens was
not found, it is interesting to note the stronger increased risk of ovarian cancer with sEPT
use and the more modest increased risk with cEPT use. This is plausible given the
biological evidence suggesting progesterone's protective role against ovarian tumor
development. If the estrogen component in HT is driving the increased ovarian cancer
56
risk seen, it is possible that the inclusion of a daily progestin counters the effects of the
estrogen and hence lowers that increased risk.
In addition to the schedule of progestin intake, the total dose of progestin during a
woman's 28-day cycle may play a role. As shown in Table 3.1, the daily dose of progestin
tends to be lower for cEPT (2.5 mg/day MPA) than for sEPT (5-10 mg/day MPA) in the
U.S. and the United Kingdom, resulting in both EPT regimens having comparable total
progestin doses. However, in some European countries, such as Sweden, the same daily
dose of progestin (1 mg/day norethisterone acetate) is used regardless of the EPT
regimen, which would result in a higher overall progestin dose for cEPT (1 mg/day of
norethisterone acetate is equivalent to 10 mg/day of MPA). The Swedish study by Riman
et al showed no increased risk with cEPT use, suggesting that the combination of daily
progestin intake at a high enough dose may block the risk-inducing effect of estrogen
(51). However, another European study by Morch et al (63) found statistically significant
increased risks for sEPT and cEPT use in their Danish cohort, despite the fact that the
hormone doses for both EPT regimens in Denmark are thought to be comparable to those
found in Sweden (64).
As shown in Table 3.4, the EPT regimen-specific results across the studies are
inconsistent and there are several possible explanations for these discrepancies. First,
each study may have defined the regimens differently. For example, in Rossing et al's
case-control study (59), having used progestin 25 days or more per month was classified
as cEPT whereas having used progestin for less than 25 days was classified as other EPT,
which I have assumed as sEPT in the meta-analysis presented. This definition is very
different from Morch et al's study (63), which defined cEPT as using the estrogen and
57
progestin components daily and sEPT as having up to seven times more estrogen than
progestin. These varying definitions would likely result in differences in exposure
classification, which could explain the differential findings. Second, inherent analytic
differences could account for some of the inconsistencies as well. For example, the MWS
(60) included only current HT users whereas Rossing et al included both current and
former users. Rossing et al also restricted their analyses to exclusive use of each EPT
regimen, which resulted in very small numbers. While this limited the precision of their
effect estimates, it also provided a cleaner analysis for determining sEPT’s and cEPT’s
independent effects. Therefore, the assessment of the EPT-ovarian cancer association will
include clarifying the associations between sEPT or cEPT use and ovarian cancer risk,
focusing specifically on the scheduling of the progestin; this is presented in Chapter 5.
Another feature worth exploring with regard to the EPT-ovarian cancer
association is whether it is characterized by histotype-specific effects. The Collaborative
Group’s pooled analysis (49) reported increased risks for serous and endometrioid
ovarian cancer with EPT use, which is similar to the findings from Morch et al’s cohort
study (54) although their data constituted a large portion of the statistical information
used by the Collaborative Group. Two additional HT studies have examined this
question, but with conflicting results; Danforth et al found no increased risk for either
histotype (52) and Moorman et al found no risk associated with serous, but decreased
risks for endometrioid and mucinous ovarian cancer (56). It is also particularly relevant to
determine whether use of sEPT and cEPT is differentially associated with risk of specific
ovarian cancer histotypes. There have only been two studies that have examined this
question to date. Riman et al found increased risks for serous and endometrioid tumors
58
with sEPT use but only for endometrioid tumors with cEPT use (51); Morch et al found
an increased risk of endometrioid tumors with sEPT use, but not with cEPT use (54). The
current literature on this subject is extremely limited, hence it will be worthwhile to
evaluate these histotype-specific differences. This will also be included in the assessment
of the EPT-ovarian cancer association presented in Chapter 5.

The potential modifying effect of genetic variation
Risk of ovarian cancer is explained by both genetic and non-genetic factors as
well as the interplay between the two. The interactive effects of six confirmed ovarian
cancer susceptibility loci identified using genome-wide association studies and five well-
established ovarian cancer risk and protective factors have previously been evaluated
with no significant interactions found (65). However, HT was not considered in these
analyses. Given the undoubted hormonal nature of ovarian cancer’s etiology, it seems
logical to evaluate potential interactions between HT and confirmed variants identified
using genome-wide association studies on risk of disease, which no study has assessed to
date.
There are currently 18 confirmed ovarian cancer susceptibility loci, each with a
relatively modest RR estimate associated with it (66-73). However, it is possible that
these associations are modified by HT use, putting some women at higher risk, which
would be important for future risk prediction modeling. Given that the existing literature
suggests ET to be an important ovarian cancer risk factor, this dissertation will focus on
gene-environment interactions specifically with regard to ET use in Chapter 6.
59
Table 3.1. Progestin Doses by EPT Regimen in the United States and Sweden (24)

Country EPT Regimen
# of Days of
Progestin
Progestin Dose
(mg/day)
Total Progestin
Dose (mg)
United States
sEPT 10 5-10 50-100
cEPT 28 2.5 70
Sweden
sEPT 10 10 100
cEPT 28 10 280
Abbreviations: EPT, estrogen-progestin combined therapy; sEPT, sequential estrogen-
progestin combined therapy; cEPT, continuous estrogen-progestin combined therapy.

60
Table 3.2. Current Literature on the Association of ET Use on Risk of Serous and Mucinous Ovarian Cancer

Study
Serous Mucinous
N RR 95% CI N RR 95% CI
Risch et al (53) 212 1.37 0.88, 2.14 40 0.72 0.24, 2.13
Purdie et al (55) 415 1.26 0.78, 2.03 114 0.85 0.32, 2.23
Riman et al (51) 337 1.33 0.85, 2.09 60 2.40 1.02, 5.65
Moorman et al (56) 216 2.00 1.30, 3.10 26 0.90 0.30, 2.50
Danforth et al (52) 233 1.25 1.12, 1.38
Morch et al (54) 1,336 1.70 1.40, 2.10 293 0.30 0.10, 0.80
Collaborative Group
(49)
2,208 1.58 1.39, 1.80 303 1.00 0.75, 1.33
Abbreviations: ET, estrogen-only therapy, RR = relative risk, CI = confidence interval.

61
Table 3.3. Current Literature on the Association of ET Use on Risk of Endometrioid and Clear Cell Ovarian Cancer

Study
Endometrioid Clear Cell
N RR 95% CI N RR 95% CI
Risch et al (53) 73 1.89 1.01, 3.54
Purdie et al (55)
a
164 2.56 1.32, 4.94 164 2.56 1.32, 4.94
Riman et al (51) 180 1.57 0.88, 2.78 43
Moorman et al (56)
a
65 1.00 0.50, 2.50 65 1.00 0.50, 2.50
Danforth et al (52) 60 1.53 1.20, 1.94
Morch et al (54) 377 1.50 1.00, 2.40 159 0.60 0.20, 1.50
Collaborative Group
(49)
508 1.34 1.05, 1.72 172 0.75 0.57, 0.98
Abbreviations: RR = relative risk, CI = confidence interval.
a
Endometrioid and clear cell cases were combined.

62
Table 3.4. Current Literature on the Association Between EPT Use and Risk of Ovarian Cancer by EPT Regimen

Study Study Type N Year Location
sEPT cEPT
RR 95% CI RR 95% CI
Riman et al (51) Case-control 4,554 2002 Sweden 1.54 1.15, 2.05 1.02 0.73, 1.43
Rossing et al (59)
a
Case-control 2,125 2006 U.S. 0.80 0.50, 1.40 0.70 0.50, 1.00
MWS (60) Cohort 948,576 2007 U.K. 1.14 0.98, 1.32 1.13 0.95, 1.33
Morch et al (63) Cohort 909,946 2009 Denmark 1.50 1.31, 1.72 1.40 1.16, 1.69
Tsildis et al (62) Cohort 126,920 2011 Europe 1.19 0.77, 1.86 1.47 0.81, 2.65
Trabert et al (61) Cohort 92,601 2012 U.S. 1.60 1.10, 2.33 1.43 1.03, 2.01
Summary 1.33 1.22, 1.45 1.18 1.06, 1.31
Abbreviations: EPT, estrogen-progestin combined therapy; sEPT, sequential estrogen-progestin combined therapy; cEPT, continuous
estrogen-progestin combined therapy; RR, relative risk; CI, confidence interval; U.S., United States; U.K., United Kingdom; MWS,
Million Women Study.
a
Considered sEPT use as <25 days of progestin use for purposes of this meta-analysis only.

63

Figure 3.1. Total Number of Adult Women Using One or More Prescribed Hormone
Therapy Drugs in the Non-Institutionalized Population of the United States, 2001 to 2008
(1)

Abbreviations: WHI, Women’s Health Initiative; HERS-II, Heart and Estrogen/Progestin
Replacement Study II.
0
2
4
6
8
10
12
14
16
18
20
2001 2002 2003 2004 2005 2006 2007 2008
Number of Women
(in millions)
Year
WHI,
HERS-II
17.9
5.8
64

Figure 3.2. Forest Plot of Study-specific and Summary RRs and 95% CIs for Risk of
Ovarian Cancer Per 5 Years of EPT Use from Pearce et al's Meta-analysis (50)

Abbreviations: RR, relative risk; CI, confidence interval; ET, estrogen-only therapy.
Study-specific RRs (boxes), summary RR (diamond), 95% CIs (lines). Overall summary
RR per 5 years of ET use (95% CI) is 1.22(1.18, 1.27), P<0.0001
65

Figure 3.3. Forest Plot of Study-specific and Summary RRs and 95% CIs for Risk of
Ovarian Cancer Per 5 Years of EPT Use from Pearce et al's Meta-analysis (50)

Abbreviations: RR, relative risk; CI, confidence interval; EPT, estrogen-progestin
combined therapy.
Study-specific RRs (boxes), summary RR (diamond), 95% CIs (lines). Overall summary
RR per 5 years of EPT use (95% CI) is 1.10 (1.04, 1.16), P=0.001.

66

REFERENCES
1. Agency for Healthcare Research and Quality. "Trends in Hormone Replacement
Therapy Drugs Utilization and Expenditures for Adult Women in the U.S.
Civilian Noninstitutionalized Population, 2001-2008". 2011.
(http://meps.ahrq.gov/mepsweb/data_files/publications/st347/stat347.pdf).
(Accessed September 20, 2013).
2. Kennedy DL, Baum C, Forbes MB. Noncontraceptive estrogens and progestins:
use patterns over time. Obstet Gynecol. 1985;65(3):441-446.
3. Smith DC, Prentice R, Thompson DJ, et al. Association of exogenous estrogen
and endometrial carcinoma. N Engl J Med. 1975;293(23):1164-1167.
4. Campbell S, McQueen J, Minardi J, et al. The modifying effect of progestogen on
the response of the post-menopausal endometrium to exogenous oestrogens.
Postgrad Med J. 1978;54 Suppl 2:59-64.
5. Walsh BW, Schiff I, Rosner B, et al. Effects of postmenopausal estrogen
replacement on the concentrations and metabolism of plasma lipoproteins. N Engl
J Med. 1991;325(17):1196-1204.
6. Miller VT, Muesing RA, LaRosa JC, et al. Effects of conjugated equine estrogen
with and without three different progestogens on lipoproteins, high-density
lipoprotein subfractions, and apolipoprotein A-I. Obstet Gynecol. 1991;77(2):235-
240.
7. Bittner V. Menopause, age, and cardiovascular risk: a complex relationship. J Am
Coll Cardiol. 2009;54(25):2374-2375.
67

8. Yang XP, Reckelhoff JF. Estrogen, hormonal replacement therapy and
cardiovascular disease. Curr Opin Nephrol Hypertens. 2011;20(2):133-138.
9. Grodstein F, Manson JE, Colditz GA, et al. A prospective, observational study of
postmenopausal hormone therapy and primary prevention of cardiovascular
disease. Ann Intern Med. 2000;133(12):933-941.
10. Grodstein F, Stampfer M. The epidemiology of coronary heart disease and
estrogen replacement in postmenopausal women. Prog Cardiovasc Dis.
1995;38(3):199-210.
11. Hersh AL, Stefanick ML, Stafford RS. National use of postmenopausal hormone
therapy: annual trends and response to recent evidence. JAMA. 2004;291(1):47-
53.
12. Nelson HD, Humphrey LL, Nygren P, et al. Postmenopausal hormone
replacement therapy: scientific review. JAMA. 2002;288(7):872-881.
13. Hulley S, Furberg C, Barrett-Connor E, et al. Noncardiovascular disease
outcomes during 6.8 years of hormone therapy: Heart and Estrogen/progestin
Replacement Study follow-up (HERS II). JAMA. 2002;288(1):58-66.
14. Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for
secondary prevention of coronary heart disease in postmenopausal women. Heart
and Estrogen/progestin Replacement Study (HERS) Research Group. JAMA.
1998;280(7):605-613.
15. Clarkson TB, Melendez GC, Appt SE. Timing hypothesis for postmenopausal
hormone therapy: its origin, current status, and future. Menopause.
2013;20(3):342-353.
68

16. Stram DO, Liu Y, Henderson KD, et al. Age-specific effects of hormone therapy
use on overall mortality and ischemic heart disease mortality among women in the
California Teachers Study. Menopause. 2011;18(3):253-261.
17. Grady D, Gebretsadik T, Kerlikowske K, et al. Hormone replacement therapy and
endometrial cancer risk: a meta-analysis. Obstet Gynecol. 1995;85(2):304-313.
18. Beral V, Bull D, Reeves G, et al. Endometrial cancer and hormone-replacement
therapy in the Million Women Study. Lancet. 2005;365(9470):1543-1551.
19. Love SM. Dr. Susan Love's Menopause and Hormone Book. New York: Three
Rivers Press; 2003.
20. Rice VM. Optimizing the dose of hormone replacement therapy. Int J Fertil
Womens Med. 2002;47(5):205-210.
21. Langer RD. Efficacy, safety, and tolerability of low-dose hormone therapy in
managing menopausal symptoms. J Am Board Fam Med. 2009;22(5):563-573.
22. Utian WH, Shoupe D, Bachmann G, et al. Relief of vasomotor symptoms and
vaginal atrophy with lower doses of conjugated equine estrogens and
medroxyprogesterone acetate. Fertil Steril. 2001;75(6):1065-1079.
23. Pike MC, Peters RK, Cozen W, et al. Estrogen-progestin replacement therapy and
endometrial cancer. J Natl Cancer Inst. 1997;89(15):1110-1116.
24. Lee SA, Ross RK, Pike MC. An overview of menopausal oestrogen-progestin
hormone therapy and breast cancer risk. Br J Cancer. 2005;92(11):2049-2058.
25. Bradley DD, Wingerd J, Petitti DB, et al. Serum high-density-lipoprotein
cholesterol in women using oral contraceptives, estrogens and progestins. N Engl
J Med. 1978;299(1):17-20.
69

26. Silfverstolpe G, Gustafson A, Samsioe G, et al. Lipid metabolic studies in
oophorectomized women. Effects of three different progestogens. Acta Obstet
Gynecol Scand Suppl. 1979;88:89-95.
27. Beral V, Reeves G, Bull D, et al. Breast cancer risk in relation to the interval
between menopause and starting hormone therapy. J Natl Cancer Inst.
2011;103(4):296-305.
28. Shah NR, Borenstein J, Dubois RW. Postmenopausal hormone therapy and breast
cancer: a systematic review and meta-analysis. Menopause. 2005;12(6):668-678.
29. Kronenberg F, Fugh-Berman A. Complementary and alternative medicine for
menopausal symptoms: a review of randomized, controlled trials. Ann Intern
Med. 2002;137(10):805-813.
30. Lukanova A, Kaaks R. Endogenous hormones and ovarian cancer: epidemiology
and current hypotheses. Cancer Epidemiol Biomarkers Prev. 2005;14(1):98-107.
31. Ho SM. Estrogen, progesterone and epithelial ovarian cancer. Reprod Biol
Endocrinol. 2003;1:73.
32. Risch HA. Hormonal etiology of epithelial ovarian cancer, with a hypothesis
concerning the role of androgens and progesterone. J Natl Cancer Inst.
1998;90(23):1774-1786.
33. Syed V, Ulinski G, Mok SC, et al. Expression of gonadotropin receptor and
growth responses to key reproductive hormones in normal and malignant human
ovarian surface epithelial cells. Cancer Res. 2001;61(18):6768-6776.
70

34. Silva EG, Tornos C, Deavers M, et al. Induction of epithelial neoplasms in the
ovaries of guinea pigs by estrogenic stimulation. Gynecol Oncol. 1998;71(2):240-
246.
35. Kurman RJ, Shih Ie M. The origin and pathogenesis of epithelial ovarian cancer:
a proposed unifying theory. Am J Surg Pathol. 2010;34(3):433-443.
36. Vang R, Shih Ie M, Kurman RJ. Fallopian tube precursors of ovarian low- and
high-grade serous neoplasms. Histopathology. 2013;62(1):44-58.
37. Olsen CM, Green AC, Whiteman DC, et al. Obesity and the risk of epithelial
ovarian cancer: a systematic review and meta-analysis. Eur J Cancer.
2007;43(4):690-709.
38. Green J, Cairns BJ, Casabonne D, et al. Height and cancer incidence in the
Million Women Study: prospective cohort, and meta-analysis of prospective
studies of height and total cancer risk. Lancet Oncol. 2011;12(8):785-794.
39. Leitzmann MF, Koebnick C, Danforth KN, et al. Body mass index and risk of
ovarian cancer. Cancer. 2009;115(4):812-822.
40. Collaborative Group on Epidemiological Studies of Ovarian C. Ovarian cancer
and body size: individual participant meta-analysis including 25,157 women with
ovarian cancer from 47 epidemiological studies. PLoS Med. 2012;9(4):e1001200.
41. Yen S. Endocrinology of Pregnancy. In: Creasy R, Resnik R, eds. Maternal-Fetal
Medicine: Principles and Practice. 3rd ed. Philadelphia, PA: Saunders; 1994:382-
412.
71

42. Johnson MR, Abbas A, Nicolaides KH. Maternal plasma levels of human
chorionic gonadotrophin, oestradiol and progesterone in multifetal pregnancies
before and after fetal reduction. J Endocrinol. 1994;143(2):309-312.
43. King RJ, Whitehead MI. Assessment of the potency of orally administered
progestins in women. Fertil Steril. 1986;46(6):1062-1066.
44. Rosenberg L, Palmer JR, Zauber AG, et al. A case-control study of oral
contraceptive use and invasive epithelial ovarian cancer. Am J Epidemiol.
1994;139(7):654-661.
45. Rodriguez GC, Walmer DK, Cline M, et al. Effect of progestin on the ovarian
epithelium of macaques: cancer prevention through apoptosis? J Soc Gynecol
Investig. 1998;5(5):271-276.
46. Barnes MN, Berry WD, Straughn JM, et al. A pilot study of ovarian cancer
chemoprevention using medroxyprogesterone acetate in an avian model of
spontaneous ovarian carcinogenesis. Gynecol Oncol. 2002;87(1):57-63.
47. Lau KM, Mok SC, Ho SM. Expression of human estrogen receptor-alpha and -
beta, progesterone receptor, and androgen receptor mRNA in normal and
malignant ovarian epithelial cells. Proc Natl Acad Sci U S A. 1999;96(10):5722-
5727.
48. Sieh W, Kobel M, Longacre TA, et al. Hormone-receptor expression and ovarian
cancer survival: an Ovarian Tumor Tissue Analysis consortium study. Lancet
Oncol. 2013;14(9):853-862.

72

49. Collaborative Group On Epidemiological Studies Of Ovarian C, Beral V,
Gaitskell K, et al. Menopausal hormone use and ovarian cancer risk: individual
participant meta-analysis of 52 epidemiological studies. Lancet.
2015;385(9980):1835-1842.
50. Pearce CL, Chung K, Pike MC, et al. Increased ovarian cancer risk associated
with menopausal estrogen therapy is reduced by adding a progestin. Cancer.
2009;115(3):531-539.
51. Riman T, Dickman PW, Nilsson S, et al. Hormone replacement therapy and the
risk of invasive epithelial ovarian cancer in Swedish women. J Natl Cancer Inst.
2002;94(7):497-504.
52. Danforth KN, Tworoger SS, Hecht JL, et al. A prospective study of
postmenopausal hormone use and ovarian cancer risk. Br J Cancer.
2007;96(1):151-156.
53. Risch HA. Estrogen replacement therapy and risk of epithelial ovarian cancer.
Gynecol Oncol. 1996;63(2):254-257.
54. Morch LS, Lokkegaard E, Andreasen AH, et al. Hormone therapy and different
ovarian cancers: a national cohort study. Am J Epidemiol. 2012;175(12):1234-
1242.
55. Purdie DM, Bain CJ, Siskind V, et al. Hormone replacement therapy and risk of
epithelial ovarian cancer. Br J Cancer. 1999;81(3):559-563.
56. Moorman PG, Schildkraut JM, Calingaert B, et al. Menopausal hormones and risk
of ovarian cancer. Am J Obstet Gynecol. 2005;193(1):76-82.
73

57. Yang HP, Anderson WF, Rosenberg PS, et al. Ovarian cancer incidence trends in
relation to changing patterns of menopausal hormone therapy use in the United
States. J Clin Oncol. 2013;31(17):2146-2151.
58. Hildebrand JS, Gapstur SM, Feigelson HS, et al. Postmenopausal hormone use
and incident ovarian cancer: Associations differ by regimen. Int J Cancer.
2010;127(12):2928-2935.
59. Rossing MA, Cushing-Haugen KL, Wicklund KG, et al. Menopausal hormone
therapy and risk of epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev.
2007;16(12):2548-2556.
60. Beral V, Million Women Study C, Bull D, et al. Ovarian cancer and hormone
replacement therapy in the Million Women Study. Lancet. 2007;369(9574):1703-
1710.
61. Trabert B, Wentzensen N, Yang HP, et al. Ovarian cancer and menopausal
hormone therapy in the NIH-AARP diet and health study. Br J Cancer.
2012;107(7):1181-1187.
62. Tsilidis KK, Allen NE, Key TJ, et al. Menopausal hormone therapy and risk of
ovarian cancer in the European prospective investigation into cancer and
nutrition. Cancer Causes Control. 2011;22(8):1075-1084.
63. Morch LS, Lokkegaard E, Andreasen AH, et al. Hormone therapy and ovarian
cancer. JAMA. 2009;302(3):298-305.
64. Lokkegaard E, Lidegaard O, Moller LN, et al. Hormone replacement therapy in
Denmark, 1995-2004. Acta Obstet Gynecol Scand. 2007;86(11):1342-1351.
74

65. Pearce CL, Rossing MA, Lee AW, et al. Combined and interactive effects of
environmental and GWAS-identified risk factors in ovarian cancer. Cancer
Epidemiol Biomarkers Prev. 2013;22(5):880-890.
66. Bolton KL, Tyrer J, Song H, et al. Common variants at 19p13 are associated with
susceptibility to ovarian cancer. Nat Genet. 2010;42(10):880-884.
67. Bojesen SE, Pooley KA, Johnatty SE, et al. Multiple independent variants at the
TERT locus are associated with telomere length and risks of breast and ovarian
cancer. Nat Genet. 2013;45(4):371-384.
68. Goode EL, Chenevix-Trench G, Song H, et al. A genome-wide association study
identifies susceptibility loci for ovarian cancer at 2q31 and 8q24. Nat Genet.
2010;42(10):874-879.
69. Kuchenbaecker KB, Ramus SJ, Tyrer J, et al. Identification of six new
susceptibility loci for invasive epithelial ovarian cancer. Nat Genet.
2015;47(2):164-171.
70. Permuth-Wey J, Kim D, Tsai YY, et al. LIN28B polymorphisms influence
susceptibility to epithelial ovarian cancer. Cancer Res. 2011;71(11):3896-3903.
71. Pharoah PD, Tsai YY, Ramus SJ, et al. GWAS meta-analysis and replication
identifies three new susceptibility loci for ovarian cancer. Nat Genet.
2013;45(4):362-370.
72. Shen H, Fridley BL, Song H, et al. Epigenetic analysis leads to identification of
HNF1B as a subtype-specific susceptibility gene for ovarian cancer. Nat
Commun. 2013;4:1628.
75

73. Song H, Ramus SJ, Tyrer J, et al. A genome-wide association study identifies a
new ovarian cancer susceptibility locus on 9p22.2. Nat Genet. 2009;41(9):996-
1000.

76
Chapter 4: Association Between Menopausal Estrogen-Only Therapy and Ovarian
Carcinoma Risk

Alice W. Lee, Roberta B. Ness, Lynda D. Roman, Kathryn L. Terry, Joellen M.
Schildkraut, Jenny Chang-Claude, Jennifer A. Doherty, Usha Menon MD, Daniel W.
Cramer, Simon A. Gayther, Harvey Risch, Aleksandra Gentry-Maharaj, Marc T.
Goodman, Francesmary Modugno, Ursula Eilber, Kirsten B. Moysich, Andrew
Berchuck, Mary Anne Rossing, Allan Jensen, Kristine G. Wicklund, Kara L. Cushing-
Haugen, Estrid Hogdall, Anja Rudolph, Pamela J. Thompson, Lynne R. Wilkens,
Susanne K. Kjaer, Michael E. Carney, Daniel O. Stram, Susan J. Ramus, Anna H. Wu,
Malcolm C. Pike, Celeste Leigh Pearce on behalf of the Ovarian Cancer Association
Consortium

77
Abbreviation List:
CI – confidence interval
ET – estrogen-only therapy
HT – hormone therapy
OC – oral contraceptive
OCAC – Ovarian Cancer Association Consortium
OR – odds ratio
WHI – Women’s Health Initiative

Running Head:
Estrogen-only Therapy and Ovarian Carcinoma Risk
78
Author Affiliations:
Department of Preventive Medicine, Keck School of Medicine, University of Southern
California, Los Angeles, California (Alice W. Lee, Daniel O. Stram, Anna H. Wu,
Malcolm C. Pike, Celeste Leigh Pearce); School of Public Health, University of Texas,
Houston, Texas (Roberta B. Ness); Department of Obstetrics and Gynecology, Keck
School of Medicine, University of Southern California, Los Angeles, California (Lynda
D. Roman); Obstetrics and Gynecology Epidemiology Center, Brigham and Women’s
Hospital, Boston, Massachusetts (Kathryn L. Terry, Daniel W. Cramer); Harvard T.H.
Chan School of Public Health, Boston, Massachusetts (Kathryn L. Terry, Daniel W.
Cramer); Department of Public Health Sciences, The University of Virginia,
Charlottesville, Virginia (Joellen M. Schildkraut); Division of Cancer Epidemiology,
German Cancer Research Center (DKFZ), Heidelberg, Germany (Jenny Chang-Claude,
Anja Rudolph); University Cancer Center Hamburg (UCCH), University Medical Center
Hamburg-Eppendorf, Hamburg, Germany (Jenny Chang-Claude);

Department of
Epidemiology, The Geisel School of Medicine at Dartmouth, Hanover, New Hampshire
(Jennifer A. Doherty); Women’s Cancer, Institute for Women’s Health, University
College London, London, United Kingdom (Usha Menon, Aleksandra Gentry-Maharaj);
Center for Cancer Prevention and Translational Genomics, Samuel Oschin
Comprehensive Cancer Institute, Cedars-Sinai Medical Center, Los Angeles, California
(Simon A. Gayther); Department of Biomedical Sciences, Cedars-Sinai Medical Center,
Los Angeles, California (Simon A. Gayther);

Department of Chronic Disease
Epidemiology, School of Public Health, Yale University, New Haven, Connecticut
(Harvey Risch); Cancer Prevention and Control, Samuel Oschin Comprehensive Cancer
79
Institute, Cedars-Sinai Medical Center, Los Angeles, California (Marc T. Goodman,
Pamela J. Thompson); Department of Biomedical Sciences, Community and Population
Health Research Institute, Cedars-Sinai Medical Center, Los Angeles, California (Marc
T. Goodman, Pamela J. Thompson); Department of Obstetrics, Gynecology, and
Reproductive Sciences, Division of Gynecologic Oncology, University of Pittsburgh
School of Medicine, Pittsburgh, Pennsylvania (Francesmary Modugno); Department of
Epidemiology, Graduate School of Public Health, University of Pittsburgh, Pittsburgh,
Pennsylvania (Francesmary Modugno);

Womens Cancer Research Program, Magee-
Womens Research Institute and University of Pittsburgh Cancer Institute, Pittsburgh,
Pennsylvania (Francesmary Modugno);

Department of Cancer Prevention and Control,
Roswell Park Cancer Institute, Buffalo, New York (Kirsten B. Moysich); Department of
Obstetrics and Gynecology, Duke University Medical Center, Durham, North Carolina
(Andrew Berchuck); Department of Epidemiology, University of Washington, Seattle,
Washington (Mary Anne Rossing); Program in Epidemiology, Division of Public Health
Sciences, Fred Hutchinson Cancer Research Center, Seattle, Washington (Mary Anne
Rossing, Kristine G. Wicklund, Kara L. Cushing-Haugen);

Department of Virus,
Lifestyle and Genes, Danish Cancer Society Research Center, Copenhagen, Denmark
(Allan Jensen, Estrid Hogdall, Susanne K. Kjaer); Department of Pathology, Molecular
Unit, Herley Hospital, University of Copenhagen, Copenhagen, Denmark (Estrid
Hogdall); Cancer Epidemiology Program, University of Hawaii Cancer Center, Honolulu,
Hawaii (Lynne R. Wilkens); Department of Gynecology, University of Copenhagen,
Copenhagen, Denmark (Susanne K. Kjaer); Department of Obstetrics and Gynecology,
John A. Burns School of Medicine, University of Hawaii, Honolulu, Hawaii (Michael E.
80
Carney); University of South Wales, Sydney, Australia (Susan J. Ramus); Department of
Epidemiology and Biostatistics, Memorial Sloan Kettering Cancer Center, New York,
New York (Malcolm C. Pike); Department of Epidemiology, School of Public Health,
University of Michigan, Ann Arbor, Michigan (Celeste Leigh Pearce)
81
To better characterize the association between postmenopausal estrogen-only
therapy (ET) use and risk of ovarian carcinoma, we conducted a pooled analysis of 906
women with ovarian carcinoma and 1,220 controls, all of whom reported having had a
hysterectomy. Ten population-based case-control studies participating in the Ovarian
Cancer Association Consortium (OCAC), an international collaboration of ovarian cancer
studies, were included. Self-reported questionnaire data from each study were
harmonized and conditional logistic regression was used to examine ET’s histotype-
specific and duration and recency of use associations. Increased risk of ovarian
carcinoma associated with current-or-recent ET use was found, but only for the serous
(odds ratio (OR)=1.63, 95% confidence interval (CI): 1.27, 2.09) and endometrioid
(OR=2.00, 95% CI: 1.17, 3.41) histotypes. In addition, statistically significant trends in
risk according to duration of use were seen among current-or-recent users for both
histotypes (P-trend<0.001

for serous and endometrioid) with current-or-recent users for
ten years or more having greater increased risks (OR=1.73, 95% CI: 1.26, 2.38 for
serous; OR=4.03, 95% CI: 1.91, 8.49 for endometrioid). These findings emphasize the
increased risk of ovarian carcinoma associated with ET use, which is characterized by
histotype-specific and duration associations.

82
Menopausal hormone therapy (HT) containing estrogens is used to relieve
climacteric symptoms and prevent osteoporosis among postmenopausal women. Prior to
the results of the Women’s Health Initiative in 2002 (1), approximately 13 million
women in the United States used HT, and while this number declined after the Women’s
Health Initiative, there are still approximately 5 million HT users (2).
A comprehensive meta-analysis by Pearce et al, which included 14 population-
based studies of women ages 18 to 79, showed that use of ET was associated with
increased risk of ovarian carcinoma (relative risk per 5 years of use=1.22) (3). Recent
studies since then have shown similar results (2, 4-6), but important aspects remain
unclear including whether differences exist by disease histotype or by duration and
timing of use. The recent pooled analysis by the Collaborative Group on Epidemiological
Studies on Ovarian Cancer (Collaborative Group) (2) did report histotype-specific
findings for serous and endometrioid cancers, but not for mucinous and clear cell cancers.
They also found little trend in association with duration of use, contrary to the results of
several studies (3, 4, 6-9). Notably, the Collaborative Group’s analysis included the
majority of studies in Pearce et al’s meta-analysis in which a duration association was
found. Clarifying these features could have important implications clinically and for risk
stratification purposes.
ET is one of the most commonly used HT types, hence a more complete
characterization of the ET-ovarian carcinoma association is warranted. We have
undertaken a pooled analysis of data from the OCAC to assess ET’s histotype-specific,
duration and recency of use associations with risk of ovarian carcinoma.

83
MATERIALS AND METHODS
The OCAC is an international multidisciplinary consortium founded in 2005
(http://apps.ccge.medschl.cam.ac.uk/consortia/ocac/index.html). Since many groups
worldwide are conducting studies to identify risk factors and genetic variation associated
with ovarian cancer risk, the goal of the OCAC is to provide a forum where data from
many individual studies with similar methods can be combined so reliable assessments of
the risks associated with these factors can be determined. Data were sent by each study
investigator to the consortium data coordinating center at Duke University, which cleaned
and harmonized these data.
For the pooled analysis presented here, ten population-based case-control studies
that were individually conducted and contributed data to the OCAC were included, with
seven conducted in the United States and three in Europe. Details regarding each study
have been published previously (10-19), but their main characteristics as well as any
overlap with the Collaborative Group’s pooled analysis are presented in Table 4.1. Cases
were women with initial diagnoses of primary ovarian carcinoma (women with primary
fallopian tube and peritoneal tumors were excluded). Eligible tumor types included
serous, mucinous, endometrioid, and clear cell ovarian carcinomas as well as other
epithelial tumor types that were not classified as one of these four main ovarian
carcinoma histotypes including mixed cell and Brenner tumors; borderline-malignant
tumors were excluded. Controls were women with ovaries (a single ovary was
acceptable), who had not been diagnosed with ovarian carcinoma at the time of interview.
Reference dates for the women in the studies were usually the dates of diagnosis for the
cases and the dates of interview for the controls. The data used in this analysis considered
84
events occurring only prior to the reference dates. All studies included in this analysis
had approval from ethics committees and written informed consent was obtained from all
study participants.
There was a total of 8,095 ovarian carcinoma patients and 13,434 controls across
the ten OCAC studies. However, only women who reported having had a simple
hysterectomy (without bilateral oophorectomy) were included in our analysis since ET
use is very infrequent among women with intact uteri as it is a confirmed risk factor for
endometrial cancer (20, 21), leaving us with 1,432 cases and 1,995 controls. Additional
exclusions included women who were less than 50 years of age at reference date (n=387),
had a prior primary cancer diagnosis (excluding non-melanoma skin cancer) (n=399), or
were missing or had unknown HT information (n=141). We also excluded women who
had used HT in an estrogen-progestin combined form (n=246) for simplicity of
presentation and since its use is likely to skew the primary effect of ET. Only women
classified as non-Hispanic white, Hispanic white, or black were considered, hence our
final subject set consisted of 2,126 women who had undergone hysterectomy, with 906
ovarian carcinoma cases and 1,220 controls (Table 4.2 and Figure 4.1).
Information regarding HT use in all forms as well as potential confounding
variables selected a priori, including age, race/ethnicity, education, oral contraceptive
(OC) use, parity, endometriosis, tubal ligation, age at menarche, and body mass index
(typically one year before the reference date), was reported by means of self-completed
questionnaires or in-person or phone interviews; we did not have information on previous
salpingectomy or BRCA status at the time of this analysis. The questions used to ascertain
HT use and, more specifically, ET use are presented in Table 4.3.
85
Age at menopause among women who have had a simple hysterectomy cannot be
determined since the women are no longer menstruating but may still have functioning
ovaries. Hence, in our primary analysis here, we have only considered ET use after age
50 given that 50 is the approximate average age at menopause for women in these
populations (22). The majority of ET use before age 50 is thus likely to be use when the
women were still having regular ovulatory cycles. Given that menopause plays a central
role in ovarian carcinoma etiology, it is possible that the added estrogen exposure during
the period when endogenous levels of estrogen are naturally high (i.e., before
menopause) is less important than exposure at older ages, the majority of which will be in
the postmenopausal period (23). Hence, for the analysis presented here, we have defined
ET use as use after age 50, with women who only used ET before age 50 included in the
baseline ‘never’ users group. We also conducted sensitivity analyses to see if the results
were affected if true ‘never’ users were used as the baseline comparison group and if ET
use was considered regardless of age at use.
A common approach to dealing with the problem of an unknown age at
menopause for women who had a hysterectomy is to use their age at simple hysterectomy
as their age at menopause. Hence, we conducted a sensitivity analysis to assess the
association between ET use and ovarian carcinoma risk using such an approach. We also
conducted sensitivity analyses using ages 48 and 52 instead of 50 as the age at
menopause.
ET use was categorized in terms of its recency and its duration of use (in years).
Current use was defined as having last used ET within the past year, recent use as within
the last one to four years, and past use as five or more years before the reference date.
86
Because current and recent ET users showed similar effects, they were combined in the
analyses presented here. Duration of ET use was summed over all episodes of use and the
total categorized into the following groups: ‘never’ (including <1 year), 1 to <5 years, 5
to <10 years, and 10 or more years of use. Women who used ET for less than one year
were included in the baseline ‘never’ users group as the recall of such short-term use may
be greater in cases than controls. All data were cleaned and checked for internal
consistency and clarifications were requested from the study investigators when needed.
Study, age, race/ethnicity, education, and OC use were included in all statistical
models. We conditioned on study, age in five-year groups (50-54, 55-59, 60-64, 65-69,
70-74, 75+; finer stratification after age 75 was not warranted due to small numbers),
race/ethnicity (non-Hispanic white, Hispanic white, and black), and education (less than
high school, high school, some college, and college graduate or higher) and we adjusted
for OC use in categories as ordinal variables ( ‘never’ (including <1 year), 1 to <2 years,
2 to <5 years, 5 to <10 years, and 10 or more years for OC use). Tubal ligation,
endometriosis, parity, body mass index, and age at menarche were also considered, but
their inclusion did not change the beta coefficients for the association between ET use
and ovarian carcinoma (including overall, serous, or endometrioid) by more than 10% so
the results given below are only adjusted for OC use. Overall, cases were missing 1.7%
and 1.1% and controls 1.4% and 0.7% for OC use and education, respectively; missing
categories were created for these women so their data could be used in the analysis.
Conditional logistic regression was used to calculate ORs and their 95% CIs for
the association between ET use and risk of ovarian carcinoma. This was done for all
ovarian carcinoma cases combined and for its four main histotypes. Similar analytic
87
approaches were applied when assessing the effects of recency and duration of use. All p-
values reported are two-sided. All analyses were performed using SAS version 9.4 (SAS
Institute, Cary, North Carolina).

RESULTS
Data from 906 women with ovarian carcinoma (567 serous, 113 endometrioid, 49
mucinous, 42 clear cell, 135 epithelial but not specified as one of the four main
histotypes) and 1,220 controls, all of whom had a simple hysterectomy, were included in
our analysis. Of these women, 460 cases (50.8%) and 531 controls (43.5%) reported ever
having used ET after age 50. Women who had used ET after age 50 had a 30% increased
risk of ovarian carcinoma as shown in Table 4.4 (OR=1.30, 95%: CI 1.06, 1.59). Most of
this risk elevation was observed among long-term users of ET for 10 years or more (both
current-or-recent and past users).
In addition, the ET-ovarian carcinoma association appeared to show distinct
histotype-specific associations as presented in Table 4.5 (serous and endometrioid) and
Table 4.6 (mucinous and clear cell). Current-or-recent ET use was statistically
significantly associated with an increased risk of both serous (OR=1.63, 95% CI: 1.27,
2.09) and endometrioid (OR=2.00, 95% CI: 1.17, 3.41) histotypes, but not mucinous
(OR=0.80, 95% CI: 0.38, 1.69) and clear cell (OR=0.80, 95% CI: 0.38, 1.68) histotypes,
although the confidence limits for the mucinous and clear cell effect estimates were wide
due to small numbers of cases. When we looked at high-grade (moderately differentiated,
poorly differentiated, undifferentiated) and low-grade (well differentiated) serous ovarian
88
carcinomas separately, we found increased risks for both and hence the results for all
serous cases combined are given.
Trends in association with duration of ET use were observed for the serous (P-
trend <0.001) and endometrioid (P-trend<0.001) histotypes among current-or-recent ET
users. Across all histotypes and duration and timing categories, ET appeared to have the
strongest association with risk of endometrioid ovarian carcinoma, with current-or-recent,
long-term users of ET for 10 years or more having over a four-fold increased risk
(OR=4.03, 95% CI: 1.91, 8.49). Current-or-recent, long-term users also had nearly a two-
fold increased risk of serous ovarian carcinoma (OR=1.73, 95% CI: 1.26, 2.38). In
addition, there appeared to be elevated risks of 1.49, 2.07, and 1.82 for overall, serous,
and endometrioid ovarian carcinoma, respectively, when we compared past, long-term
ET users to our baseline ‘never’ user group (Tables 4.4 and 4.5).
Because we assumed that all women in our analysis had an age at menopause of
50, we conducted a sensitivity analysis in which each woman’s age at simple
hysterectomy was used as her age at menopause, with the duration and timing of use
variables re-categorized as such. The results by duration, timing of ET use, and histotype
slightly attenuated with ORs of 1.46, 1.64, and 3.72 among current-or-recent ET users of
10 years or more for ovarian carcinoma overall and the serous and endometrioid
histotypes, respectively (Table 4.7). Sensitivity analyses that used a true ‘never’ user
baseline group and redefined ET use regardless of age at menopause or with ages 48 and
52 as the age at menopause did not affect the overall findings (data not shown).

89
DISCUSSION
Most population-based case-control studies and cohort studies have shown that
ET use is associated with an increased risk of ovarian carcinoma and considering our
findings together with those recently published by the Collaborative Group (2), it seems
clear that ET is associated with risk for the serous and endometrioid histotypes of the
disease. We found greater increased risk for those who used ET for 10 years or more,
including those who last used it more than five years in the past, whereas the
Collaborative Group (2) did not. This was surprising given that the individual studies that
contributed the most statistical information to their analysis (the Million Women Study
(7) and the Danish Sex Hormone Register Study (4)) reported duration associations with
ET use in their primary publications. The meta-analysis from Pearce and colleagues (3)
showed evidence of an ET duration-ovarian carcinoma risk association as well.
From a biological standpoint, an elevated risk of endometrioid ovarian carcinoma
with ET use is not surprising given that the cells of origin are histologically similar to
endometrial tissue (24), and ET use is a confirmed risk factor for endometrial cancer (20,
21). Danforth et al had suggested that ET may act through similar biologic mechanisms
in the development of endometrioid tumors as it does in endometrial cancer (25). Given
the increased risk we see for endometrioid ovarian carcinoma and the well-established
association between endometriosis and the endometrioid and clear cell histotypes (26),
we assessed the ET risk association according to previous history of endometriosis or not,
but did not see any heterogeneity in risk (data not shown).
Although the exact mechanism by which ET might affect serous and
endometrioid ovarian carcinoma risk remains unknown, estrogens have long been
90
implicated as etiologic factors (27). Ovarian carcinogenesis may be a result of the direct
effects of unopposed estrogen and an estrogen-rich environment, which would potentially
be enhanced by ET use. The use of ET may also directly stimulate the growth of
premalignant or early malignant cells with long-term use increasing the risk of
transformation or proliferation (28). In addition, the fallopian tube fimbriae, a proposed
cell of origin for high-grade serous cancer, have been shown to proliferate at times when
estrogenic influences are greater during the menstrual cycle (29, 30), and this increased
activity results in greater cell proliferation which may enhance the risk of mutations and
malignant transformation. Estradiol has also been shown to increase ovarian carcinoma
cell proliferation in vitro (31) and influence the growth of ovarian tumors in a
transplanted mouse model (32). Therefore, while several hypotheses have been put forth
to explain ovarian carcinoma etiology, unopposed estrogen appears to play an important
role.
Limitations of our analysis include the self-reported nature of our data. Because
case-control studies inquire about previous exposures when subjects are already aware of
their disease status, recall bias is possible as cases may be more likely to search for
explanations for their disease and assign greater significance to past events than controls.
However, studies have shown high agreement between self-reported estrogen use and
prescription data (33). In addition, case patients have not been shown to preferentially
report HT use more than controls (34). We considered ET use only after age 50 to be
relevant in an attempt to mainly consider only use after ovarian function had ceased.
Sensitivity analysis showed little effect when changing this to age 48 or 52, the latter
which will only include use that is almost all in the postmenopausal period.
91
A potential concern with case-control studies such as those included in our
analysis is that some ineligible women (those who had a bilateral oophorectomy) could
have been recruited as controls even though they would not be at risk of developing
ovarian carcinoma. However, oophorectomy results in a loss of estrogen production,
which may make such women more likely to use ET, thus potentially biasing our findings
towards the null. If this type of bias is present, any association between ET use and risk
of ovarian carcinoma would be underestimated.
Our analysis offers evidence of an increased risk of ovarian carcinoma with ET
use after the age of 50. This is especially true for risk of serous and endometrioid tumors
for long durations of use, shedding light on the distinct histotype-specific etiologies.
Although ET use has declined since the WHI, a significant number of women continue to
use it today. Physicians and patients should be aware of the risk of ovarian carcinoma
associated with its long-term use.

92

Table 4.1. Characteristics of Studies Included in Analysis

Study Name
a

Time
Period
Location
Case
Ascertainment
Control
Ascertainment
Connecticut Ovary Study (17) 2002-2009 Connecticut
Cancer registry or hospital
records
Random digit dialing, Health
Care Financing Administration
records
Disease of the Ovary and Their
Evaluation Study (16)
2002-2009 Washington Cancer registry Random digit dialing
German Ovarian Cancer Study
(10)
b

1992-1998 Germany
Admissions to all hospitals
serving the study regions
Population registries
Hawaii Ovarian Cancer Study
(12)
1994-2007 Hawaii Cancer registry
Department of Health Annual
Survey, Health Care Financing
Administration records
Hormones and Ovarian Cancer
Prediction (13)
b

2003-2008
Western
Pennsylvania,
Northeast Ohio,
Western New
York
Cancer registries, pathology
databases, physicians’
offices
Random digit dialing
Malignant Ovarian Cancer Study
(18)
1994-1999 Denmark
Cancer registry,
gynecological departments
Random digit dialing
North Carolina Ovarian Cancer
Study (14)
1999-2008 North Carolina Cancer registry Random digit dialing
New England Case-Control
Study of Ovarian Cancer (15)
1999-2008
New
Hampshire,
Eastern
Massachusetts
Cancer registries, hospital
tumor boards
Random digit dialing, town
books, drivers’ license lists
United Kingdom Ovarian Cancer
Population Study (11)
2006-2007
United
Kingdom
Gynecological Oncology
National Health Service
centers
Women from the United
Kingdom Collaborative Trial of
Ovarian Cancer Screening

93

Table 4.1., Continued

Study Name
a

Time
Period
Location
Case
Ascertainment
Control
Ascertainment
University of Southern
California, Study of Lifestyle
and Women’s Health (19)
b

1993-2005
Los Angeles,
California
Cancer registry Neighborhood controls
a
All studies used in-person interviews except the Germany Ovarian Cancer Study, which used self-completed questionnaires.
b
Some of the study’s data were included in the Collaborative Group’s analysis (2).
94

Table 4.2. Numbers of Cases and Controls Included in Analysis by Study

Study Name
Controls
(Mean/IQR
for Age)
Cases
(Mean/IQR
for Age)
Serous Mucinous
Endom-
etrioid
Clear
Cell
Connecticut Ovary Study (17) 49 (60.9/13) 54 (62.2/12) 28 4 12 4
Disease of the Ovary and Their
Evaluation Study (16)
224 (66.8/12) 159 (62.7/8) 108 3 15 3
German Ovarian Cancer Study
(10)
89 (60.9/11) 34 (61.5/12) 17 2 2 1
Hawaii Ovarian Cancer Study
(12)
40 (67.3/16) 32 (66.1/17) 18 1 5 2
Hormones and Ovarian Cancer
Prediction (13)
201 (65.5/16) 100 (66.6/15.6) 57 5 17 4
Malignant Ovarian Cancer Study
(18)
84 (62.6/13) 47 (61.7/11) 25 4 8 5
North Carolina Ovarian Cancer
Study (14)
126 (62.7/11) 153 (62.9/10) 94 5 16 11
New England Case-Control Study
of Ovarian Cancer (15)
67 (63.2/11) 50 (63.4/12) 38 1 6 1
United Kingdom Ovarian Cancer
Population Study (11)
116 (64.0/9) 56 (67.0/12) 30 7 12 4
University of Southern
California. Study of Lifestyle and
Women’s Health (19)
224 (62.9/12.5) 217 (64.7/12) 152 17 20 7
Total 1,220 906 567
a
49
a
113
a
42
a

Abbreviations: IQR, interquartile range.
a
Sum of numbers do not equal total number of cases because some cases were not classified as one of the four main histotypes
considered.
95
Table 4.3. Hormone Therapy Questionnaire Questions by Study

Study Name Questions
Connecticut Ovary
Study (17)
Have you ever used estrogen patches or pills for menopause?
Here is a card with possible medications you might have used.
(If yes) What were they, how old were you when you started,
and how long did you use them for?
Disease of the Ovary
and their Evaluation
Study (16)
Before (reference date), did you ever use any hormone
medications just before the onset of menopause, around the
time of menopause, or after menopause? What is the name of
the hormone you were using? (with photobook)
German Ovarian
Cancer Study (10)
Have you ever taken other hormones (e.g. estrogen, gestagen,
progesterone etc.)? What was the hormone preparation,
reason, from (year) duration? (provides sample list)
Hawaii Ovarian
Cancer Study (12)
Please tell me if you were ever given hormones before
(reference date) for each of the following reasons. What was
the name and dose of the hormone you used?
Hormones and
Ovarian Cancer
Prediction (13)
Before (reference date), have you ever taken medications for
menopause etc. (list of reasons)? What was the name of the
hormone pill you took from this card? (with photo cards)
Malignant Ovarian
Cancer Study (18)
Did you ever receive hormone therapy? (ie. hormonal
treatment of symptoms prior to or after menopause, but not
hormonal treatment of infertility problems). In what form did
you have this hormone therapy? What brand did you use?
North Carolina
Ovarian Cancer Study
(14)
Have you ever used estrogens, progestins, or other male or
female hormones (other than for birth control or for breast
cancer)? Have you ever taken Premarin, Prempro, Premphase
or other conjugated estrogens? Have you ever taken other
estrogen pills? Have you ever used estrogen patches?
New England Case-
Control Study of
Ovarian Cancer (15)
Do you remember using any kind of hormonal medications to
treat symptoms that you or your doctor thought were due to
the menopause, or to prevent problems like osteoporosis due
to the menopause? Thinking about the hormonal preparation
you used to treat menopausal symptoms, do you recall using
estrogen pills alone? Estrogen patch alone? What was the
name of the pill or patch?
United Kingdom
Ovarian Cancer
Population Study (11)
Have you ever used hormone replacement therapy (HRT)?
Please name the HRT you used (you could check the names
from the list the nurse has with her) and at what age you
started. Estimate the number of years/months you used the
preparation.
University of Southern
California, Study of
Lifestyle and
Women's Health (19)
Please tell me if you were ever given hormones before
(reference date). What was the name and dose of the hormone
you used?
96
Table 4.4. Association Between ET Use Over Age 50 and Risk of Ovarian Carcinoma
Overall

Number
of
Controls
Number
of Cases
Median
Duration
(years)
OR
a
95% CI P Value
Never used 689 446 1.00
Ever 531 460 9.20 1.30 1.06, 1.59 0.013
*

1 to <5 years 149 92 2.70 1.00 0.72, 1.39 0.99
5 to <10 years 155 135 7.45 1.27 0.93, 1.72 0.13
10+ years 227 233 15.12 1.54 1.18, 2.01 0.002
**

P-trend 0.001
***

Current-or-
recent users
432 392 10.00 1.35 1.09, 1.67 0.006
***

1 to <5 years 103 67 3.00 1.00 0.68, 1.48 0.99
5 to <10 years 120 112 7.20 1.35 0.96, 1.90 0.087
10+ years 209 213 15.20 1.53 1.17, 2.02 0.002
***

P-trend <0.001
***

Past users 99 68 6.20 1.07 0.74, 1.56 0.72
1 to <5 years 46 25 2.20 1.01 0.59, 1.74 0.97
5 to <10 years 35 23 8.20 1.03 0.57, 1.86 0.93
10+ years 18 20 13.28 1.49 0.71, 3.13 0.29
P-trend 0.95
Abbreviations: ET, estrogen-only therapy; OR, odds ratio; CI, confidence interval.
*
P≤0.05,
**
P≤0.01,
***
P≤0.001
a
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+
years) and conditioned on age (50-54, 55-59, 60-64, 65-69, 70-74, 75+), education (less
than high school, high school, some college, college graduate or higher), race/ethnicity
(non-Hispanic white, Hispanic white, black), and study.

97
Table 4.5. Association Between ET Use After Age 50 and Risk of Serous and Endometrioid Ovarian Carcinoma

Serous Endometrioid
Number
of
Controls
Number
of Cases
OR
a
95% CI P Value
Number
of cases
OR
a
95% CI P Value
Never used 689 252 1.00 54 1.00
Ever 531 315 1.57 1.23, 2.00 <0.001
***
59 1.82 1.10, 3.03 0.021
1 to <5 years 149 62 1.26 0.86, 1.83 0.24 10 0.98 0.45, 2.15 0.97
5 to <10 years 155 92 1.58 1.11, 2.25 0.012
*
17 1.64 0.78, 3.47 0.19
10+ years 227 161 1.79 1.31, 2.43 <0.001
***
32 3.58 1.74, 7.36 <0.001
***

P-trend: <0.001
***
<0.001
***

Current-or-
recent users
432 267 1.63 1.27, 2.09 <0.001
***
51 2.00 1.17, 3.41 0.011
*

1 to <5 years 103 46 1.37 0.88, 2.14 0.16 7 0.88 0.35, 2.19 0.78
5 to <10 years 120 74 1.69 1.14, 2.52 0.010
**
15 1.72 0.76, 3.87 0.19
10+ years 209 147 1.73 1.26, 2.38 <0.001
***
29 4.03 1.91, 8.49 <0.001
***

P-trend: <0.001
***
<0.001
***

Past users 99 48 1.28 0.83, 1.96 0.27 8 1.20 0.48, 3.01 0.69
1 to <5 years 46 16 1.05 0.55, 2.01 0.89 3 1.35 0.37, 4.94 0.66
5 to <10 years 35 18 1.26 0.65, 2.43 0.50 2 1.46 0.25, 8.66 0.68
10+ years 18 14 2.07 0.89, 4.79 0.091 3 1.82 0.40, 8.19 0.44
P-trend: 0.46 0.35
Abbreviations: ET, estrogen-only therapy; OR, odds ratio; CI; confidence interval.
*
P≤0.05,
**
P≤0.01,
***
P≤0.001
a
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+ years) and conditioned on age (50-54, 55-
59, 60-64, 65-69, 70-74, 75+), education (less than high school, high school, some college, college graduate or higher), race/ethnicity
(non-Hispanic white, Hispanic white, black), and study.

98
Table 4.6. Association Between ET Use After Age 50 and Risk of Mucinous and Clear Cell Ovarian Carcinoma

Mucinous Clear Cell

Number
of
controls
Number
of cases
OR
a
95% CI
P
value
Number
of cases
OR
a
95% CI
P
value
Never used 689 33 1.00 25 1.00
Ever 531 16 0.80 0.38, 1.69 0.57 17 0.80 0.38, 1.68 0.55
1 to <5 years 149 2 0.28 0.06, 1.43 0.13 5 1.21 0.40, 3.71 0.74
5 to <10 years 155 6 1.33 0.44, 3.98 0.61 6 1.04 0.34, 3.13 0.95
10+ years 227 8 0.92 0.33, 2.55 0.88 6 0.51 0.17, 1.50 0.22
P-trend: 0.98 0.31

Current-or-
recent users
335 15 0.93 0.43, 2.00 0.84 16 0.87 0.40, 1.88 0.72
1 to <5 years 80 2 0.43 0.08, 2.42 0.34 5 1.84 0.55, 6.23 0.32
5 to <10 years 85 5 1.19 0.36, 3.92 0.78 6 1.39 0.44, 4.35 0.58
10+ years 170 8 0.99 0.36, 2.72 0.98 5 0.40 0.13, 1.29 0.12
P-trend: 0.99 0.25

Past users 196 1 0.30 0.04, 2.40 0.26 1 0.42 0.05, 3.49 0.42
1 to <5 years 69 0 0
5 to <10 years 70 1 2.06 0.18, 23.45 0.56 0
10+ years 57 0 1 1.67 0.13, 22.05 0.70
P-trend:
Abbreviations: ET, estrogen-only therapy; OR, odds ratio; CI, confidence interval.
a
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+ years) and conditioned on age (50-54, 55-
59, 60-64, 65-69, 70-74, 75+), education (less than high school, high school, some college, college graduate or higher), race/ethnicity
(non-Hispanic white, Hispanic white, black), and study.

99

Table 4.7. Association Between Current-or-Recent ET use and Risk of Ovarian
Carcinoma by Histotype Using Age at Hysterectomy as Age at Menopause

Number
of
controls
Number
of cases
OR
a
95% CI P Value
Overall Invasive
Never used 609 402 1.00
Current-or-recent users 451 402 1.28 1.03, 1.59 0.024
*

1 to <5 years 76 54 1.00 0.67, 1.48 0.98
5 to <10 years 106 80 1.31 0.93, 1.85 0.13
10+ years 269 268 1.46 1.10, 1.93 0.009
**

P-trend: 0.003
**

Serous
Never used 609 227 1.00
Current-or-recent users 451 271 1.54 1.20, 1.99 <0.001
1 to <5 years 76 35 1.34 0.85, 2.11 0.21
5 to <10 years 106 56 1.64 1.09, 2.44 0.017
*

10+ years 269 180 1.64 1.19, 2.27 0.003
**

P-trend: <0.001
***

Endometrioid
Never used 609 48 1.00
Current-or-recent users 451 54 1.84 1.08, 3.12 0.025
*

1 to <5 years 76 6 0.79 0.30, 2.08 0.63
5 to <10 years 106 11 1.56 0.68, 3.59 0.29
10+ years 269 37 3.72 1.72, 8.02 <0.001
***

P-trend: 0.002
**

Abbreviations: ET, estrogen-only therapy; OR, odds ratio; CI, confidence interval.
*
P≤0.05,
**
P≤0.01,
***
P≤0.001
a
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+
years) and conditioned on age (50-54, 55-59, 60-64, 65-69, 70-74, 75+), education (less
than high school, high school, some college, college graduate and higher), race/ethnicity
(non-Hispanic white, Hispanic white, black), and study.

100

Figure 4.1. Flowchart of Analysis Exclusions

101

REFERENCES
1. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus
progestin in healthy postmenopausal women: principal results From the Women's
Health Initiative randomized controlled trial. JAMA. 2002;288(3):321-333.
2. Collaborative Group On Epidemiological Studies Of Ovarian C, Beral V,
Gaitskell K, et al. Menopausal hormone use and ovarian cancer risk: individual
participant meta-analysis of 52 epidemiological studies. Lancet.
2015;385(9980):1835-1842.
3. Pearce CL, Chung K, Pike MC, et al. Increased ovarian cancer risk associated
with menopausal estrogen therapy is reduced by adding a progestin. Cancer.
2009;115(3):531-539.
4. Morch LS, Lokkegaard E, Andreasen AH, et al. Hormone therapy and ovarian
cancer. JAMA. 2009;302(3):298-305.
5. Tsilidis KK, Allen NE, Key TJ, et al. Menopausal hormone therapy and risk of
ovarian cancer in the European prospective investigation into cancer and
nutrition. Cancer Causes Control. 2011;22(8):1075-1084.
6. Hildebrand JS, Gapstur SM, Feigelson HS, et al. Postmenopausal hormone use
and incident ovarian cancer: Associations differ by regimen. Int J Cancer.
2010;127(12):2928-2935.
7. Beral V, Million Women Study C, Bull D, et al. Ovarian cancer and hormone
replacement therapy in the Million Women Study. Lancet. 2007;369(9574):1703-
1710.
102

8. Riman T, Dickman PW, Nilsson S, et al. Hormone replacement therapy and the
risk of invasive epithelial ovarian cancer in Swedish women. J Natl Cancer Inst.
2002;94(7):497-504.
9. Moorman PG, Schildkraut JM, Calingaert B, et al. Menopausal hormones and risk
of ovarian cancer. Am J Obstet Gynecol. 2005;193(1):76-82.
10. Royar J, Becher H, Chang-Claude J. Low-dose oral contraceptives: protective
effect on ovarian cancer risk. Int J Cancer. 2001;95(6):370-374.
11. Balogun N, Gentry-Maharaj A, Wozniak EL, et al. Recruitment of newly
diagnosed ovarian cancer patients proved challenging in a multicentre biobanking
study. J Clin Epidemiol. 2011;64(5):525-530.
12. Lurie G, Terry KL, Wilkens LR, et al. Pooled analysis of the association of
PTGS2 rs5275 polymorphism and NSAID use with invasive ovarian carcinoma
risk. Cancer Causes Control. 2010;21(10):1731-1741.
13. Ness RB, Dodge RC, Edwards RP, et al. Contraception methods, beyond oral
contraceptives and tubal ligation, and risk of ovarian cancer. Ann Epidemiol.
2011;21(3):188-196.
14. Moorman PG, Calingaert B, Palmieri RT, et al. Hormonal risk factors for ovarian
cancer in premenopausal and postmenopausal women. Am J Epidemiol.
2008;167(9):1059-1069.
15. Terry KL, De Vivo I, Titus-Ernstoff L, et al. Androgen receptor cytosine, adenine,
guanine repeats, and haplotypes in relation to ovarian cancer risk. Cancer Res.
2005;65(13):5974-5981.
103

16. Bodelon C, Cushing-Haugen KL, Wicklund KG, et al. Sun exposure and risk of
epithelial ovarian cancer. Cancer Causes Control. 2012;23(12):1985-1994.
17. Risch HA, Bale AE, Beck PA, et al. PGR +331 A/G and increased risk of
epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2006;15(9):1738-
1741.
18. Glud E, Kjaer SK, Thomsen BL, et al. Hormone therapy and the impact of
estrogen intake on the risk of ovarian cancer. Arch Intern Med.
2004;164(20):2253-2259.
19. Wu AH, Pearce CL, Tseng CC, et al. African Americans and Hispanics Remain at
Lower Risk of Ovarian Cancer Than Non-Hispanic Whites after Considering
Nongenetic Risk Factors and Oophorectomy Rates. Cancer Epidemiol
Biomarkers Prev. 2015;24(7):1094-1100.
20. Beral V, Bull D, Reeves G, et al. Endometrial cancer and hormone-replacement
therapy in the Million Women Study. Lancet. 2005;365(9470):1543-1551.
21. Grady D, Gebretsadik T, Kerlikowske K, et al. Hormone replacement therapy and
endometrial cancer risk: a meta-analysis. Obstet Gynecol. 1995;85(2):304-313.
22. Morabia A, Costanza MC. International variability in ages at menarche, first
livebirth, and menopause. World Health Organization Collaborative Study of
Neoplasia and Steroid Contraceptives. Am J Epidemiol. 1998;148(12):1195-1205.
23. Pike MC. Age-related factors in cancers of the breast, ovary, and endometrium. J
Chronic Dis. 1987;40 Suppl 2:59S-69S.
24. Kumar V CR, Robbins SL. Basic Pathology. 6 ed. Philadelphia, PA: W.B.
Saunders; 1997.
104

25. Danforth KN, Tworoger SS, Hecht JL, et al. A prospective study of
postmenopausal hormone use and ovarian cancer risk. Br J Cancer.
2007;96(1):151-156.
26. Pearce CL, Templeman C, Rossing MA, et al. Association between endometriosis
and risk of histological subtypes of ovarian cancer: a pooled analysis of case-
control studies. Lancet Oncol. 2012;13(4):385-394.
27. Ho SM. Estrogen, progesterone and epithelial ovarian cancer. Reprod Biol
Endocrinol. 2003;1:73.
28. Zhou H, Luo MP, Schonthal AH, et al. Effect of reproductive hormones on
ovarian epithelial tumors: I. Effect on cell cycle activity. Cancer Biol Ther.
2002;1(3):300-306.
29. Donnez J, Casanas-Roux F, Caprasse J, et al. Cyclic changes in ciliation, cell
height, and mitotic activity in human tubal epithelium during reproductive life.
Fertil Steril. 1985;43(4):554-559.
30. George SH, Milea A, Shaw PA. Proliferation in the normal FTE is a hallmark of
the follicular phase, not BRCA mutation status. Clin Cancer Res.
2012;18(22):6199-6207.
31. Nash JD, Ozols RF, Smyth JF, et al. Estrogen and anti-estrogen effects on the
growth of human epithelial ovarian cancer in vitro. Obstet Gynecol.
1989;73(6):1009-1016.
32. Laviolette LA, Hodgkinson KM, Minhas N, et al. 17beta-estradiol upregulates
GREB1 and accelerates ovarian tumor progression in vivo. Int J Cancer. 2014.
105

33. Sandini L, Pentti K, Tuppurainen M, et al. Agreement of self-reported estrogen
use with prescription data: an analysis of women from the Kuopio Osteoporosis
Risk Factor and Prevention Study. Menopause. 2008;15(2):282-289.
34. Paganini-Hill A, Clark LJ. Comparison of patient recall of hormone therapy with
physician records. Menopause. 2007;14(2):230-234.

106
Chapter 5: The Association Between Menopausal Estrogen-Progestin Therapy and
Risk of Ovarian Cancer: A Pooled Analysis

107
Abbreviation List:
cEPT – continuous estrogen-progestin combined therapy
CI – confidence interval
EPT – estrogen-progestin combined therapy
ET – estrogen-only therapy
HT – hormone therapy
OC – oral contraceptive
OCAC – Ovarian Cancer Association Consortium
OR – odds ratio
RR – relative risk
sEPT – sequential estrogen-progestin combined therapy

Running Head:
Estrogen-progestin Therapy and Ovarian Cancer Risk

108
Menopausal hormone therapy (HT) is used for climacteric symptom relief and
osteoporosis prevention. Its estrogen-only therapy (ET) form is a confirmed risk factor
for endometrial cancer, but with concurrent progestin use (estrogen-progestin combined
therapy (EPT)), estrogen’s carcinogenic effects on the endometrium are mitigated. Some
studies have shown progestin’s protective effect to extend to the ovary, but results have
been inconsistent. We pooled primary data from 11 population-based case-control studies
participating in the Ovarian Cancer Association Consortium (OCAC), including 3,265
postmenopausal ovarian cancer cases and 5,519 controls. Conditional logistic regression
was used to assess EPT’s effect by duration, recency, type of regimen as well as by
disease histotype. We found EPT to be associated with a 13% decreased risk of ovarian
cancer (odds ratio (OR)=0.87, 95% confidence interval (CI): 0.77, 0.97). This was largely
driven by mucinous (OR=0.41, 95% CI: 0.25, 0.67) and clear cell tumors (OR=0.53, 95%
CI: 0.37, 0.76) and those who used EPT continuously (OR=0.80, 95% CI: 0.68, 0.94).
Interestingly, current EPT users of 10+ years showed a 33% increased risk (OR=1.33,
95% CI: 1.04, 1.70), which diminished after use ceased. Overall, these findings highlight
how EPT’s effect must be considered in light of both exposure and disease
characteristics.

109
HT supplements the lack of naturally occurring estrogen and progesterone after
menopause and is one of the most effective treatments for menopausal symptoms.
However, the results of the Women’s Health Initiative randomized trial in 2002 showed
its use, particularly EPT, to be associated with an increased risk of breast cancer (1).
Despite the Women’s Health Initiative report, a significant number of women in the
United States have continued using HT with over five million users currently (2).
ET is a confirmed risk factor for endometrial cancer, but the addition of a
progestin (ie. EPT) counteracts its carcinogenic effect on the endometrium (3). ET is also
a risk factor for ovarian cancer and there is some evidence to suggest that adding a
progestin ameliorates this effect. Laboratory studies have shown progesterone to have
apoptotic effects and inhibitory action on ovarian cancer cells (4). Also, several
epidemiologic studies have shown that the increased risk of ovarian cancer associated
with use of ET is largely attenuated when a progestin is used concurrently (5-7),
including a comprehensive meta-analysis conducted by Pearce et al (8). However, the
most recent analysis by the Collaborative Group on Epidemiological Studies of Ovarian
Cancer (Collaborative Group) did not report such an attenuation (9).
Clarifying the association between EPT and risk of ovarian cancer may
significantly contribute to our understanding of the effects of female sex hormones on the
ovary. Thus, we have a conducted a pooled analysis using primary data from the Ovarian
OCAC to assess whether EPT is associated with risk of ovarian cancer, with particular
attention paid to duration and timing of use, type of EPT regimen, and disease histotype.

110
MATERIALS AND METHODS
All studies included in this report obtained institutional ethics committee
approval. All participating subjects provided written informed consent.

Study populations
Our analysis included data from 11 population-based case-control studies
participating in the OCAC, an international collaboration of ovarian cancer studies
(http://apps.ccge.medschl.cam.ac.uk/consortia/ocac/index.html). Specifics regarding each
individual study have been published previously (10-22), but their main characteristics
and any overlap with the Collaborative Group’s analysis (9) are summarized in Table 5.1.
Cases were women with primary first diagnoses of invasive epithelial ovarian
cancer; women with primary fallopian tube and peritoneal tumors were excluded. Serous,
mucinous, endometrioid, clear cell as well as other invasive epithelial ovarian cancers
that were not classified as one of these four histotypes were considered in our analyses.
Controls were women who reported having at least one intact ovary and had not been
diagnosed with ovarian cancer at or before the date of interview.
Across the 11 OCAC studies, there was a total of 14,948 postmenopausal women
(5,710 ovarian cancer cases and 9,238 controls). Only women who were non-Hispanic
white, Hispanic white, or black and had no prior history of primary cancers (excluding
non-melanoma skin cancer) were included in our analyses, leaving us with 4,603 cases
and 7,691 controls. In addition, we excluded women who had missing or unknown HT
information (226 cases and 469 controls), women who used EPT for less than one year
(261 cases and 470 controls) as well as women who reported prior use of ET (1,042 cases
111
and 1,585 controls) since ET is a well-established ovarian cancer risk factor (8, 9, 23) and
its prior use may mask EPT’s true association with the disease. Hence, our final subject
set included 3,097 ovarian cancer cases and 5,195 controls. Figure 5.1 presents a
flowchart of these exclusions.

Exposure and covariate information
All data pertaining to HT use as well as potential confounding variables were self-
reported using self-completed questionnaires, phone interviews, or in-person interviews.
Life calendars and photobooks were used to aid memory when necessary. The questions
used to ascertain HT and, more specifically, EPT exposure are presented in Table 5.2.
The collected questionnaire data considered only HT use prior to each woman’s reference
date (date of diagnosis for cases, date of interview for controls). Age, race/ethnicity,
education, oral contraceptive (OC) use, parity, hysterectomy, endometriosis, tubal
ligation, age at menopause, and body mass index (typically one year before the woman’s
reference date) were selected a priori as potential confounders.
Because menopause plays an important role in ovarian cancer etiology and is
characterized by major changes in endogenous hormone levels, the effects of EPT when
used before menopause could differ from its effects when used after menopause when
hormone levels are naturally lower (24). However, premenopausal and postmenopausal
EPT use appeared to be associated with little difference in disease risk so EPT was
considered regardless of its use relative to menopause (see Sensitivity Analyses below).
Use of EPT was categorized according to recency (current, recent, and past) and duration
of use. We classified current users as those who used EPT within the last year, recent
112
users as those who used EPT within the last one to four years, and past users as those
who last used EPT five or more years before their reference dates. However, because
recent and past users showed similar effects that were distinct from current users, the two
categories were combined (from hereinafter referred to as “past users”). Duration of use
was determined by summing all episodes of EPT use with the total duration categorized
as never use, 1 to <5 years, 5 to <10 years, and 10+ years of use. Those who used EPT
for <1 year were not considered because the effects are likely to be minimal and recall of
such short-term use may be greater in cases than controls.
In addition, we further categorized EPT use according to its two main regimens:
sequential estrogen-progestin combined therapy (sEPT) and continuous estrogen-
progestin combined therapy (cEPT). This categorization was based on the average
number of days per month the progestin was used concurrently with the estrogen. We
classified sEPT users as those who used a progestin with daily ET on average for 5 to 15
days per month and cEPT users as those who used a progestin with daily ET for 25 days
or more per month. Progestin use for <5 days or for 16 to 24 days was classified as
neither, and hence women with these durations were not included in these subgroup
analyses.

Statistical analysis
To estimate the main effects of EPT use on risk of ovarian cancer, ORs, as
estimates of the relative risk (RR), and 95% CIs were calculated using conditional
logistic regression. All models were conditioned on study and age in five-year categories
(<40, 40-44, 45-49, 50-54, 55-59, 60-64, 65-69, 70-74, 75+). We also adjusted for
113
race/ethnicity (non-Hispanic white, Hispanic white, black), education (less than high
school, high school graduate, some college, college graduate and higher), duration of OC
use (never use (including <1 year), 1 to <2 years, 2 to <5 years, 5 to <10 years, 10+
years), body mass index (<18.5, 18.5-24.9, 25-29.9, 30+ kg/m
2
), hysterectomy status
(yes/no), and tubal ligation (yes/no). Personal history of endometriosis (yes/no) and
parity (0, 1, 2+ births) were considered, but their inclusion did not change the beta
coefficients for the association between EPT use and ovarian cancer by more than 10%,
hence the models presented are without their adjustments. The same approach was used
for analyses evaluating duration and timing of EPT use as well as disease histotype.
Seven of the 11 OCAC studies included asked HT questions such that use of
sEPT and cEPT could be derived (Table 5.2). The sEPT and cEPT subgroup analyses
were carried out as described above, however only ever/never use with respect to risk
overall and by histotype were evaluated due to the small sample size available. We also
restricted these analyses to only those who used one type of EPT regimen in order to
assess sEPT’s and cEPT’s independent effects.

Sensitivity analyses
To compare the effects of premenopausal versus postmenopausal EPT use, use of
EPT was re-categorized such that only women who used it for at least one year after their
reported age at menopause were considered postmenopausal EPT users. These women
were compared to a baseline ‘never’ user group that included those who never used HT,
those who used EPT for less than one year after menopause, and those who only used
EPT premenopausally; we also used a true ‘never’ user group for these analyses and did
114
not see any difference in the results. Since we saw little difference with regard to EPT’s
effect, the results we present look at EPT use overall.
In addition, because the analyses on the effects of sEPT and cEPT use were only
done on the seven studies that had information on type of EPT regimen used, sensitivity
analyses were conducted to evaluate the EPT-ovarian cancer association among this
subset of studies for comparison purposes. We did not observe any difference in the
overall results as well (data not shown).

All tests of statistical significance were two-sided. The analyses were performed
using SAS software, release 9.4 (SAS Institute, Inc., Cary, North Carolina).

RESULTS
Our analyses included a total of 3,097 ovarian cancer cases (1,862 serous, 177
mucinous, 425 endometrioid, 245 clear cell, 388 invasive, epithelial but not classified as
one of the four main histotypes) and 5,195 controls (Table 5.1). Among these women,
755 cases (24.3%) and 1,466 controls (28.2%) reported using EPT for at least one year.
Overall, use of EPT was associated with a 13% decreased risk of ovarian cancer
(OR=0.87, 95% CI: 0.77, 0.97) (Table 5.3). Long-term EPT use for 10+ years was not
associated with ovarian cancer risk (OR=1.01, 95% CI: 0.84, 1.21), but after considering
recency of use, we noted an increased risk among current, long-term users (OR=1.33,
95% CI: 1.04, 1.70) and a decreased risk among past, long-term users (OR=0.82, 95%
CI: 0.64, 1.04) with effect estimates that were statistically significantly different from one
another (P-heterogeneity=0.006).
115
In addition, the association between EPT use and risk of ovarian cancer appeared
to be characterized by histotype-specific differences (Table 5.4). EPT use was associated
with decreased risk for both mucinous (OR=0.41, 95% CI: 0.25, 0.67) and clear cell
(OR=0.53, 95% CI: 0.37, 0.76) histotypes, and this protective effect was seen across all
duration categories assessed. There appeared to be no association between EPT use and
the serous and endometrioid histotypes, regardless of duration (Table 5.4), but when this
was further evaluated by recency of use, increased risk was observed among current EPT
users of 10+ years for both (OR=1.58, 95% CI: 1.20, 2.08 for serous, OR=1.38, 95% CI:
0.81, 2.35 for endometrioid). Current versus past use was not evaluated for the mucinous
and clear cell histotypes due to limited numbers. The serous histotype was evaluated
according to tumor grade (high-grade – moderately differentiated, poorly differentiated,
undifferentiated; low-grade – well-differentiated), but there was little difference in risk,
hence, the results for all serous cases combined are presented.
Table 5.5 presents the results for ovarian cancer overall according to use of sEPT
and cEPT regimens from the seven studies that provided this information. Statistical
power for this analysis is limited, but for all cases combined, cEPT was associated with a
decreased risk of ovarian cancer (OR=0.80, 95% CI: 0.68, 0.94). This protective effect
for cEPT was also consistent across all histotypes, though only mucinous and clear cell
reached statistical significance. Conversely, there was no association with sEPT use and
overall ovarian cancer risk (OR=0.96, 95% CI: 0.74, 1.25). The histotype-specific results
showed a borderline statistically significant reduced risk of mucinous ovarian cancer and
a statistically significant reduced risk of clear cell ovarian cancer for sEPT use, but this
116
result was based on three and two cases using sEPT, respectively; there was no
association between sEPT use and risk of the serous or endometrioid histotypes.
Interestingly, while the EPT analyses for risk of ovarian cancer showed an overall
protective association as presented in Table 5.3, detailed analyses of duration of use
among all cases combined indicate that this decreased risk was predominantly driven by
cEPT users (Table 5.6). Long-term cEPT users of 10+ years showed a 23% decreased
risk of ovarian cancer overall (OR=0.77, 95% CI: 0.58, 1.03) whereas sEPT users in the
same duration category showed a 27% increased risk (OR=1.27, 95% CI: 0.81, 1.99),
albeit neither effect estimate was statistically significant.

DISCUSSION
A better understanding of the effects of EPT use on ovarian cancer risk is needed
given its importance among postmenopausal women with intact uteri. Our pooled
analysis showed that overall EPT use appeared to be associated with a slight decreased
risk of ovarian cancer. However, this association was characterized by risk differences
according to type of EPT regimen used as well as disease histotype, highlighting the
complexity of ovarian cancer’s etiology.
In contrast to our findings, the Collaborative Group found a significant increased
risk of ovarian cancer among current EPT users, which is surprising given the evidence
supporting the hypothesized protective role of progesterone in ovarian carcinogenesis in
both epidemiologic studies (25, 26) and experimental studies (4, 27). While estrogens
have been implicated as causative factors for ovarian cancer development as
demonstrated by the well-established effects of ET (9, 23, 28), it has been proposed that
117
the addition of a progestin may diminish estrogens’ risk-inducing effects (8), which is in
line with our results. Concurrent use of a progestin is needed with ET in order to
counteract the carcinogenic effects of unopposed estrogen on the endometrium (3). Since
recent evidence has suggested the fallopian tube fimbria as the cell of origin for serous
ovarian cancer (29) and studies of cell proliferation in the fallopian tube have shown that
it follows the same pattern in relation to the menstrual cycle as the endometrium (30), the
progestin may play a similar role in both ovarian and endometrial carcinogenesis.
While we observed an overall protective association between EPT use and risk of
ovarian cancer, there appeared to be differences when recency and duration of use are
considered. Current users of EPT for 10+ years showed an increased risk for ovarian
cancer overall and the serous and endometrioid histotypes, but this increased risk
diminished after EPT use ceased. Given the proposed biological roles of estrogen and
progesterone, it is possible that the progestin component still confers a risk reduction, but
that this effect is masked by the opposing risk-inducing effect of estrogen, which persists
until EPT use has ceased as suggested by Rossing et al who also found a protective effect
among former EPT users of extended durations (5). However, few studies have evaluated
the EPT-ovarian cancer association by both recency and duration of use so these findings
should be interpreted with caution. To our knowledge, only three studies have examined
this association in such detail with inconsistent results (5, 6, 31).
In addition, our results suggest that the frequency of concurrent progestin use with
estrogen differentially impacts disease risk; this has also been observed for both
endometrial cancer (32) and breast cancer (33). The overall protective effect we observe
for the EPT-ovarian cancer association appeared to be driven by cEPT use as sEPT use
118
showed little risk except when used for extended periods of time. Although statistical
power was limited, these observations do support the notion that estrogen’s effect on the
ovary may be offset by progestin. Hence, daily use of progestin (ie. cEPT) could have a
greater risk-reducing effect than a regimen in which the progestin is only taken on certain
days (ie. sEPT). However, the current literature has been inconsistent (7, 31, 34), which
may be due to varying definitions of sEPT and cEPT as well as different analytic
approaches.
The decreased risk observed with EPT use was primarily driven by mucinous and
clear cell tumors, similar to the results of the Collaborative Group’s analysis (9).
Mucinous tumors have been thought to be gastrointestinal in origin, and studies have
shown HT use to be associated with reduced risks of esophageal, gastric, and colorectal
cancer (35, 36). Interestingly, the clear cell histotype, like endometrioid, is
endometriosis-associated, but the effect of HT on these respective tumors appears to be
quite distinct; ET use does not appear to be associated with clear cell ovarian cancer, but
has been found to increase risk of endometrioid ovarian cancer (9, 23). This may be
attributable to differences in hormone receptor expression (37) or may simply reflect
inherent etiologic differences (38).
Most studies to date have not evaluated the EPT-ovarian cancer association in
such detail due to limited numbers. However, given our use of primary data from 11
population-based case-control studies and the detailed nature of each study’s
questionnaire, we have been able to consider several exposure and disease features. One
limitation to our analysis is the self-reported nature of our primary data. However, efforts
were taken to aid in the recall of HT use, including the use of photobooks and life
119
calendars. In addition, studies have shown high concordance between self-reported HT
data and prescription records with little indication that disease status would differentially
impact recall of HT use (39, 40).
Although the Collaborative Group reported a 25% increased risk of ovarian
cancer with current EPT use (RR=1.25, 95% CI: 1.16, 1.34), their supplemental analyses
by study design indicate that this increased risk was primarily driven by the prospective
studies they included (9). In fact, they did not find an association between current EPT
use and risk of ovarian cancer among retrospective studies (RR=0.96, 95% CI: 0.83,
1.10), which is more similar to our results (9). Despite the strengths associated with
prospective studies, they often lack detailed exposure data. For example, in the
Collaborative Group’s analysis, each woman’s HT status was based on what she last
used, hence previous use of ET would be disregarded if the woman last switched to an
EPT prescription. Similarly, a woman may have last used ET, but due to lack of follow-
up data, may have been classified as having last used EPT. These scenarios could be
problematic given the well-established increased risk of ovarian cancer associated with
ET use. Although use of EPT is unlikely to follow use of ET since only hysterectomized
women should use ET to begin with, this does not appear to be the case in our study
population as a little more than half the women who used both HT types reported using
ET prior to EPT. Hence, our pooled analysis only considered exclusive use of EPT.
In addition, the Collaborative Group could have more sEPT than cEPT users,
especially in their prospective studies which appear to be driving their EPT results.
Although they do not report on EPT regimen-specific effects, the two studies that
contributed the most statistical information to their analysis of prospective studies
120
(Million Women Study (7) and the Danish Sex Hormone Register Study (34)) did have
greater numbers of cases who were sEPT users versus cEPT users. In addition, studies
have shown that those of higher socioeconomic status (SES) are more likely to participate
in prospective studies (41) and within our own data, women who were more educated
were more likely to use sEPT; the Collaborative Group did not consider SES in their
analyses.
In conclusion, our findings highlight the complexity of ovarian carcinogenesis
with regard to estrogen and progesterone. We have shown that while EPT use is
associated with a decreased risk of ovarian cancer, this protection must be considered in
light of both exposure and disease characteristics as the relationship appears to be driven
by different effects by histotype, type of regimen (sEPT versus cEPT), and duration and
recency of use. Our results shed light on the roles female sex hormones play in ovarian
cancer etiology and more importantly, have important clinical and public health
implications. While EPT use may be more advantageous with respect to ovarian cancer
risk, this must be weighed against the impact of this regimen on breast cancer risk.
Women, in conjunction with their physician, must consider the totality of risks and
benefits associated with HT use when deciding whether and which type to use.

121
Table 5.1. Numbers of Cases and Controls Included in Analysis by Study

Study Name Location
Study
Period
Controls Cases Serous
Muci-
nous
Endom-
etrioid
Clear
Cell
Connecticut Ovary Study (19)
a
USA; CT 2002-2009 206 169 98 5 32 21
Disease of the Ovary and Their
Evaluation Study (18)
USA; WA 2002-2009 619 369 229 7 51 30
German Ovarian Cancer Study (10)
a,b
Germany 1992-1998 245 111 58 12 7 3
Hawaii Ovarian Cancer Study (11) USA; HI 1994-2007 134 112 62 7 13 4
Hormones and Ovarian Cancer
Prediction (12)
b

USA; northeast
OH, western
PA and NY
2003-2008 695 261 144 9 38 20
Malignant Ovarian Cancer Study (21) Denmark 1994-1999 682 273 167 30 35 19
North Carolina Ovarian Cancer Study
(13)
a

USA; NC 1999-2008 423 373 202 14 57 48
New England Case-Control Study of
Ovarian Cancer (17)
USA; NH and
eastern MA
1999-2008 450 322 211 17 51 18
New Jersey Ovarian Cancer Study
(20)
USA; NJ 2002-2008 244 98 55 5 14 16
United Kingdom Ovarian Cancer
Population Study (22)
a

United
Kingdom
2006-2007 557 268 140 21 48 26
University of Southern California,
Study of Lifestyle and Women’s
Health (15)
b

USA; Los
Angeles, CA
1993-2005 940 741 496 50 79 40
Total 5,195 3,097 1,862
c
177
c
425
c
245
c

Abbreviations: EPT, estrogen-progestin combined therapy.
a
Excluded from sequential EPT and continuous EPT analyses due to lack of EPT regimen-specific information.
b
Some of the study’s primary data were included in the Collaborative Group’s (9) analysis.
c
Sum of numbers do not equal total number of cases because some cases were not classified as one of the four histotypes considered.
122

Table 5.2. Hormone Therapy Questionnaire Questions by Study

Study Name Questions
Connecticut
Ovary Study (19)
a

Have you ever used estrogen patches or pills for menopause?
What were they, how old were you when you started, and how
long did you use them for?
Disease of the
Ovary and Their
Evaluation Study
(18)
Before (reference date), did you ever use any hormone
medications just before the onset of menopause, around the time
of menopause, or after menopause?
What is the name of the hormone you were using?
When did you first use this hormone?
During a month when you were using (hormone), did you usually
take it every day of the month or in cycles – that is, what was your
pattern of use?
German Ovarian
Cancer Study
(10)
a

Have you ever taken other hormones (estrogen, gestagen,
progesterone etc.)?
What was the hormone preparation, reason, from (year), duration?
Hawaii Ovarian
Cancer Study
(11)
Please tell me if you were ever given hormones before (reference
date) near the time of or after menopause.
Were they in the form of pills, shots, vaginal creams, something
else?
When did you start/stop using hormone pills?
What was the name and dose of the hormone you used? How
many days per month did you use it?
Did you use any other hormones at the same time? What was the
name and dose of that hormone? How many days per month did
you use it?
Hormones and
Ovarian Cancer
Prediction (12)
Before (reference date), have you ever taken medications for
menopause?
Did you ever take any of these pills either alone or in combination
with a progesterone (such as Provera) for a period of at least one
month? Remember these are hormones used for reasons other than
birth control.
What was the name of the hormone pill you took? What was the
dose?
What was the name of the progestin pill you took with the
estrogen? What was the dose?
North Carolina
Ovarian Cancer
Study (13)
a

Have you ever used estrogens, progestins, or other male or female
hormones such as Danazol (other than for birth control or breast
cancer)?
What brand and dose did you use?
What was your age at first/last use?
Overall, for how many months or years did you use (hormone)?

123

Table 5.2., Continued

Study Name Questions
New England
Case-Control
Study of Ovarian
Cancer (17)
Do you remember using any hormonal preparations to treat
symptoms that you or your doctor thought were due to menopause,
or to prevent problems like osteoporosis due to menopause?
Thinking about the hormonal preparation you used to treat
menopausal symptoms, do you recall using estrogen pills alone,
estrogen patch alone, estrogen pills or patch combined with
progesterone etc.?
What was the name of the pill or patch? What age did you first use
it? How many months total did you use it?
If yes to used estrogen combined with progesterone, was the
progesterone taken daily with the estrogen? Taken only 10-14
days each month?
New Jersey
Ovarian Cancer
Study (20)
Before (reference date), did you ever use hormones to relieve
symptoms of menopause?
What was the name of the hormone you used?
What was the dose of the hormone you used?
In what month and year did you first use/stop using hormone pills?
Did you use any other hormones at the same time? If yes, what
was the name of the hormone you used? How many days per
month did you use it?
United Kingdom
Ovarian Cancer
Population Study
(22)
a

Have you ever taken hormone replacement therapy?
Please name the hormone therapy you used (you could check the
names from the list the nurse has with her) and at what age you
started and estimate the number of years/months you used the
preparation.
University of
Southern
California, Study
of Lifestyle and
Women's Health
(15)
Please tell me if you were ever given hormones before (reference
date) near the time of or after menopause to alleviate menopausal
symptoms, to treat menopausal depression etc.
When did you start using hormone pills? When did you stop using
them or change dosage or change your pattern of use?
What was the name and dose of the hormone you used? How
many days per month did you use it?
Did you use any other hormones at the same time? If yes, name
and dose? How many days per month did you use it?
Abbreviations: EPT, estrogen-progestin combined therapy.
a
Excluded from EPT regimen-specific analyses since the questionnaire questions could
not properly differentiate between sequential EPT and continuous EPT use.

124

Table 5.3. Association Between EPT Use and Risk of Ovarian Cancer Overall

Number
of
Controls
a

Number
of
Cases
a

Median
Duration
(years)
OR
a
95% CI P value
Never used 3,729 2,342 -- 1.00
Ever 1,466 755 6.36 0.87 0.77, 0.97 0.013
*

1-4.99 years 578 258 2.44 0.74 0.63, 0.88 <0.001
***

5-9.99 years 464 245 7.28 0.89 0.75, 1.06 0.21
10+ years 424 252 13.36 1.01 0.84, 1.21 0.92
P-trend 0.33

Current users 511 307 8.00 0.95 0.80, 1.12 0.51
1 to <5 years 177 80 2.36 0.71 0.54, 0.95 0.022
*

5 to <10 years 151 86 7.00 0.86 0.65, 1.14 0.30
10+ years 183 131 13.36 1.33 1.04, 1.70 0.025
*

P-trend 0.31

Past users 955 448 6.36 0.82 0.72, 0.94 0.004
**

1 to <5 years 401 178 2.53 0.76 0.63, 0.93 0.007
**

5 to <10 years 313 159 7.36 0.93 0.75, 1.14 0.47
10+ years 241 111 13.53 0.82 0.64, 1.04 0.10
P-trend 0.011
*

Abbreviations: EPT, estrogen-progestin combined therapy; OR, odds ratio; CI,
confidence interval.
*
P≤0.05,
**
P≤0.01,
***
P≤0.001
a
Adjusted for education (less than high school, high school, some college, college
graduate and higher), race/ethnicity (non-Hispanic white, Hispanic white, black), oral
contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+ years,
hysterectomy (yes/no), tubal ligation (yes/no), and body mass index (<18.5, 18.5-24.9,
25-29.9, 30+ kg/m
2
); conditioned on age (<40, 40-44, 45-49, 50-54, 55-59, 60-64, 65-69,
70-74, 75+) and study.

125

Table 5.4. Association Between EPT Use and Risk of Ovarian Cancer by Histotype

Number
of
Controls
Number
of Cases
OR
a
95% CI P value
Serous
Never used 3,729 1,357 1.00
Ever used 1,466 505 0.97 0.85, 1.11 0.63
1 to <5 years 578 171 0.85 0.70, 1.03 0.098
5 to <10 years 464 160 0.97 0.79, 1.19 0.74
10+ years 424 174 1.14 0.93, 1.40 0.21
P-trend 0.54
Endometrioid
Never used 3,729 315 1.00
Ever used 1,466 110 0.99 0.77, 1.27 0.94
1 to <5 years 578 38 0.82 0.57, 1.19 0.30
5 to <10 years 464 36 1.07 0.73, 1.57 0.72
10+ years 424 36 1.16 0.79, 1.71 0.46
P-trend 0.55
Mucinous
Never used 3,729 158 1.00
Ever used 1,466 19 0.41 0.25, 0.67 <0.001
***

1 to <5 years 578 4 0.20 0.07, 0.54 0.002
**

5 to <10 years 464 8 0.55 0.26, 1.16 0.12
10+ years 424 7 0.59 0.27, 1.31 0.20
P-trend 0.011
*

Clear Cell
Never used 3,729 201 1.00
Ever used 1,466 44 0.53 0.37, 0.76 <0.001
***

1 to <5 years 578 16 0.49 0.29, 0.84 0.010
***

5 to <10 years 464 17 0.63 0.37, 1.07 0.089
10+ years 424 11 0.48 0.25, 0.91 0.025
*

P-trend 0.002
**

Abbreviations: EPT, estrogen-progestin combined therapy; OR, odds ratio, CI,
confidence interval.
*
P≤0.05,
**
P≤0.01,
***
P≤0.001
a
Adjusted for education (less than high school, high school, some college, college
graduate and higher), race/ethnicity (non-Hispanic white, Hispanic white, black), oral
contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+ years),
hysterectomy (yes/no), tubal ligation (yes/no), and body mass index (<18.5, 18.5-24.9,
25-29.9, 30+); conditioned on age (<40, 40-44, 45-49, 50-54, 55-59, 60-64, 65-69, 70-74,
75+) and study.

126
Table 5.5. Association Between sEPT Use and cEPT Use and Risk of Ovarian Cancer by Histotype

sEPT cEPT
Number
of
Controls
a

Number
of
Cases
a

OR
b
95% CI P value
Number
of
Controls
a

Number
of
Cases
a

OR
b
95% CI P value
Overall Invasive
Never used 2,745 1,679 1.00 2,745 1,679 1.00
Ever used 196 101 0.96 0.74, 1.25 0.77 726 328 0.80 0.68, 0.94 0.006
**

Serous
Never used 2,745 1022 1.00 2,745 1,022 1.00
Ever used
†
196 74 1.10 0.82, 1.48 0.53 726 220 0.86 0.71, 1.03 0.10
Endometrioid
Never used 2,745 211 1.00 2,745 211 1.00
Ever used 196 15 1.06 0.60, 1.90 0.84 726 46 0.98 0.68, 1.41 0.90
Mucinous
Never used 2,745 115 1.00 2,745 115 1.00
Ever used 196 3 0.37 0.12, 1.22 0.10 726 6 0.30 0.13, 0.70 0.005
Clear Cell
Never used 2,745 122 1.00 2,745 122 1.00
Ever used 196 2 0.24 0.06, 0.98 0.046
*
726 19 0.59 0.35, 1.00 0.050
*

Abbreviations: sEPT, sequential estrogen-progestin combined therapy; cEPT, continuous estrogen-progestin combined therapy; OR,
odds ratio; CI, confidence interval.
*
P≤0.05,
**
P≤0.01
a
Numbers only include women from the following studies: DOV, HAW, HOP, MAL, NEC, NJO, USC.
b
Adjusted for education (less than high school, high school, some college, college graduate and higher), race/ethnicity (non-Hispanic
white, Hispanic white, black), oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+ years), hysterectomy
(yes/no), tubal ligation (yes/no), and body mass index (<18.5, 18.5-24.9, 25-29.9, 30+); conditioned on age (<40, 40-44, 45-49, 50-54,
55-59, 60-64, 65-69, 70-74, 75+) and study.
127
Table 5.6. Association Between sEPT Use and cEPT Use and Risk of Ovarian Cancer Overall by Duration

sEPT cEPT
Number
of
Controls
a

Number
of
Cases
a

OR
b
95% CI P value
Number
of
Controls
a

Number
of
Cases
a

OR
b
95% CI P value
Never used 2745 1679 1.00 2745 1679 1.00
Ever used 196 101 0.96 0.74, 1.25 0.77 726 328 0.80 0.68 – 0.94 0.006
1 to <5 years 82 36 0.75 0.49, 1.14 0.18 304 140 0.80 0.64 – 1.01 0.054
5 to <10 years 53 31 0.99 0.62 – 1.58 0.96 240 109 0.81 0.63 – 1.04 0.10
10+ years 61 34 1.27 0.81 – 1.99 0.30 182 79 0.77 0.58 – 1.03 0.079
P-trend 0.65 0.011
Abbreviations: sEPT, sequential estrogen-progestin combined therapy; cEPT, continuous estrogen-progestin combined therapy; OR,
odds ratio; CI, confidence interval.
a
Numbers only include women from the following studies: DOV, HAW, HOP, MAL, NEC, NJO, USC.
b
Adjusted for education (less than high school, high school, some college, college graduate and higher), race/ethnicity (non-Hispanic
white, Hispanic white, black), oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+ years), hysterectomy
(yes/no), tubal ligation (yes/no), and body mass index (<18.5, 18.5-24.9, 25-29.9, 30+); conditioned on age (<40, 40-44, 45-49, 50-54,
55-59, 60-64, 65-69, 70-74, 75+) and study.

128

Figure 5.1. Flowchart of Analysis Exclusions

129

REFERENCES
1. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus
progestin in healthy postmenopausal women: principal results From the Women's
Health Initiative randomized controlled trial. JAMA. 2002;288(3):321-333.
2. Hersh AL, Stefanick ML, Stafford RS. National use of postmenopausal hormone
therapy: annual trends and response to recent evidence. JAMA. 2004;291(1):47-
53.
3. Beral V, Bull D, Reeves G, et al. Endometrial cancer and hormone-replacement
therapy in the Million Women Study. Lancet. 2005;365(9470):1543-1551.
4. Yu S, Lee M, Shin S, et al. Apoptosis induced by progesterone in human ovarian
cancer cell line SNU-840. J Cell Biochem. 2001;82(3):445-451.
5. Rossing MA, Cushing-Haugen KL, Wicklund KG, et al. Menopausal hormone
therapy and risk of epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev.
2007;16(12):2548-2556.
6. Hildebrand JS, Gapstur SM, Feigelson HS, et al. Postmenopausal hormone use
and incident ovarian cancer: Associations differ by regimen. Int J Cancer.
2010;127(12):2928-2935.
7. Beral V, Million Women Study C, Bull D, et al. Ovarian cancer and hormone
replacement therapy in the Million Women Study. Lancet. 2007;369(9574):1703-
1710.
8. Pearce CL, Chung K, Pike MC, et al. Increased ovarian cancer risk associated
with menopausal estrogen therapy is reduced by adding a progestin. Cancer.
2009;115(3):531-539.
130

9. Collaborative Group On Epidemiological Studies Of Ovarian C, Beral V,
Gaitskell K, et al. Menopausal hormone use and ovarian cancer risk: individual
participant meta-analysis of 52 epidemiological studies. Lancet.
2015;385(9980):1835-1842.
10. Royar J, Becher H, Chang-Claude J. Low-dose oral contraceptives: protective
effect on ovarian cancer risk. Int J Cancer. 2001;95(6):370-374.
11. Lurie G, Terry KL, Wilkens LR, et al. Pooled analysis of the association of
PTGS2 rs5275 polymorphism and NSAID use with invasive ovarian carcinoma
risk. Cancer Causes Control. 2010;21(10):1731-1741.
12. Ness RB, Dodge RC, Edwards RP, et al. Contraception methods, beyond oral
contraceptives and tubal ligation, and risk of ovarian cancer. Ann Epidemiol.
2011;21(3):188-196.
13. Moorman PG, Calingaert B, Palmieri RT, et al. Hormonal risk factors for ovarian
cancer in premenopausal and postmenopausal women. Am J Epidemiol.
2008;167(9):1059-1069.
14. Wu AH, Pearce CL, Tseng CC, et al. Markers of inflammation and risk of ovarian
cancer in Los Angeles County. Int J Cancer. 2009;124(6):1409-1415.
15. Wu AH, Pearce CL, Tseng CC, et al. African Americans and Hispanics Remain at
Lower Risk of Ovarian Cancer Than Non-Hispanic Whites after Considering
Nongenetic Risk Factors and Oophorectomy Rates. Cancer Epidemiol
Biomarkers Prev. 2015;24(7):1094-1100.
131

16. Pike MC, Pearce CL, Peters R, et al. Hormonal factors and the risk of invasive
ovarian cancer: a population-based case-control study. Fertil Steril.
2004;82(1):186-195.
17. Terry KL, De Vivo I, Titus-Ernstoff L, et al. Androgen receptor cytosine, adenine,
guanine repeats, and haplotypes in relation to ovarian cancer risk. Cancer Res.
2005;65(13):5974-5981.
18. Bodelon C, Cushing-Haugen KL, Wicklund KG, et al. Sun exposure and risk of
epithelial ovarian cancer. Cancer Causes Control. 2012;23(12):1985-1994.
19. Risch HA, Bale AE, Beck PA, et al. PGR +331 A/G and increased risk of
epithelial ovarian cancer. Cancer Epidemiol Biomarkers Prev. 2006;15(9):1738-
1741.
20. Bandera EV, King M, Chandran U, et al. Phytoestrogen consumption from foods
and supplements and epithelial ovarian cancer risk: a population-based case
control study. BMC Womens Health. 2011;11:40.
21. Glud E, Kjaer SK, Thomsen BL, et al. Hormone therapy and the impact of
estrogen intake on the risk of ovarian cancer. Arch Intern Med.
2004;164(20):2253-2259.
22. Balogun N, Gentry-Maharaj A, Wozniak EL, et al. Recruitment of newly
diagnosed ovarian cancer patients proved challenging in a multicentre biobanking
study. J Clin Epidemiol. 2011;64(5):525-530.
23. Lee AW, Ness RB, Roman LD, et al. Association between menopausal estrogen-
only therapy and ovarian carcinoma risk. Obstet Gynecol. In press.
132

24. Pike MC. Age-related factors in cancers of the breast, ovary, and endometrium. J
Chronic Dis. 1987;40 Suppl 2:59S-69S.
25. Yen S. Endocrinology of Pregnancy. In: Creasy R, Resnik R, eds. Maternal-Fetal
Medicine: Principles and Practice. 3rd ed. Philadelphia, PA: Saunders; 1994:382-
412.
26. Risch HA. Hormonal etiology of epithelial ovarian cancer, with a hypothesis
concerning the role of androgens and progesterone. J Natl Cancer Inst.
1998;90(23):1774-1786.
27. Rodriguez GC, Walmer DK, Cline M, et al. Effect of progestin on the ovarian
epithelium of macaques: cancer prevention through apoptosis? J Soc Gynecol
Investig. 1998;5(5):271-276.
28. Ho SM. Estrogen, progesterone and epithelial ovarian cancer. Reprod Biol
Endocrinol. 2003;1:73.
29. Crum CP, Drapkin R, Miron A, et al. The distal fallopian tube: a new model for
pelvic serous carcinogenesis. Curr Opin Obstet Gynecol. 2007;19(1):3-9.
30. Donnez J, Casanas-Roux F, Caprasse J, et al. Cyclic changes in ciliation, cell
height, and mitotic activity in human tubal epithelium during reproductive life.
Fertil Steril. 1985;43(4):554-559.
31. Riman T, Dickman PW, Nilsson S, et al. Hormone replacement therapy and the
risk of invasive epithelial ovarian cancer in Swedish women. J Natl Cancer Inst.
2002;94(7):497-504.
32. Pike MC, Peters RK, Cozen W, et al. Estrogen-progestin replacement therapy and
endometrial cancer. J Natl Cancer Inst. 1997;89(15):1110-1116.
133

33. Lee SA, Ross RK, Pike MC. An overview of menopausal oestrogen-progestin
hormone therapy and breast cancer risk. Br J Cancer. 2005;92(11):2049-2058.
34. Morch LS, Lokkegaard E, Andreasen AH, et al. Hormone therapy and ovarian
cancer. JAMA. 2009;302(3):298-305.
35. Green J, Czanner G, Reeves G, et al. Menopausal hormone therapy and risk of
gastrointestinal cancer: nested case-control study within a prospective cohort, and
meta-analysis. Int J Cancer. 2012;130(10):2387-2396.
36. Lindblad M, Garcia Rodriguez LA, Chandanos E, et al. Hormone replacement
therapy and risks of oesophageal and gastric adenocarcinomas. Br J Cancer.
2006;94(1):136-141.
37. Sieh W, Kobel M, Longacre TA, et al. Hormone-receptor expression and ovarian
cancer survival: an Ovarian Tumor Tissue Analysis consortium study. Lancet
Oncol. 2013;14(9):853-862.
38. Nagle CM, Olsen CM, Webb PM, et al. Endometrioid and clear cell ovarian
cancers: a comparative analysis of risk factors. Eur J Cancer. 2008;44(16):2477-
2484.
39. Sandini L, Pentti K, Tuppurainen M, et al. Agreement of self-reported estrogen
use with prescription data: an analysis of women from the Kuopio Osteoporosis
Risk Factor and Prevention Study. Menopause. 2008;15(2):282-289.
40. Paganini-Hill A, Clark LJ. Comparison of patient recall of hormone therapy with
physician records. Menopause. 2007;14(2):230-234.
134

41. Drivsholm T, Eplov LF, Davidsen M, et al. Representativeness in population-
based studies: a detailed description of non-response in a Danish cohort study.
Scand J Public Health. 2006;34(6):623-631.

135
Chapter 6: A Splicing Variant of TERT Identified by GWAS Interacts with
Menopausal Estrogen Therapy in Risk of Ovarian Cancer

136
Abbreviation List:
CI – confidence interval
ET – estrogen-only therapy
GWAS – genome-wide association study
HT – hormone therapy
OC – oral contraceptive
OCAC – Ovarian Cancer Association Consortium
OR – odds ratio
SNP – single nucleotide polymorphism

Running Head:
Gene-Hormone Therapy Interaction in Ovarian Cancer

137
Menopausal estrogen-only therapy (ET) is a well-established risk factor for serous
and endometrioid ovarian cancer. Genetics also plays a role in ovarian cancer, which is
partly attributable to 18 confirmed ovarian cancer susceptibility loci identified by
genome-wide association studies (GWASs). The interplay between these loci and ET use
on ovarian cancer risk has yet to be evaluated. We analyzed data from 1,414 serous cases,
337 endometrioid cases, and 4,051 controls across 10 case-control studies participating in
the Ovarian Cancer Association Consortium (OCAC). Conditional logistic regression was
used to determine the association between the variants and risk of each histotype among
ET users and non-users separately and to test for statistical interaction. A splicing variant
in TERT, rs10069690, showed a statistically significant interaction with ET use for serous
ovarian cancer (P-interaction=0.013). ET users carrying the T allele showed a 48%
increased risk (odds ratio (OR)=1.48, 95% confidence interval (CI): 1.15, 1.91) whereas
non-users showed essentially no association (OR=1.09, 95% CI: 0.97, 1.22). Two regions
harboring rs7207826 (C allele) and rs56318008 (T allele) also had significant interactions
with ET use for the endometrioid histotype (P-interaction =0.021 and 0.037,
respectively). These findings, if validated, may elucidate the mechanism of action of
these loci.
138
The etiology of ovarian cancer is influenced by several hormonal factors,
including menopausal hormone therapy (HT) use. Approximately five million women in
the United States currently use HT, and according to the National Health and Nutrition
Examination Survey in 2010, the most commonly used type of HT among women aged
40 years and older is ET (1, 2). ET is a well-established risk factor for serous and
endometrioid ovarian cancer (2-4). Most recently, Lee et al demonstrated that use of ET
postmenopausally was associated with a 57% and 82% increased risk of serous and
endometrioid ovarian cancer, respectively (5); the meta-analysis by the Collaborative
Group on Epidemiological Studies of Ovarian Cancer showed these histotype effects as
well (2).
Ovarian cancer also has a strong genetic component. A large part is attributable to
high-penetrance susceptibility mutations, but common variants identified using GWASs
play important roles as well. There are currently 18 confirmed ovarian cancer common
susceptibility loci that explain approximately 3.9% of the disease’s excess familial risk
(6-13). Each of these common variants is associated with extremely modest relative risk
estimates, but it is possible that interactions between non-genetic and genetic risk factors
exist, thereby putting some women at higher risk.
Pearce et al previously examined the interactive effects between six GWAS-
identified common variants and five well-accepted non-genetic risk factors: first-degree
family history of ovarian cancer, tubal ligation, parity, oral contraceptive (OC) use, and
personal history of endometriosis (14). However, menopausal ET, which has consistently
been shown to be associated with risk of serous and endometrioid ovarian cancer (2, 5),
was not included in these analyses. Using data from the OCAC, we have evaluated
139
potential statistical interactions between menopausal ET use and the 18 confirmed
ovarian cancer common susceptibility alleles. To our knowledge, this is the first study to
investigate the interactions between menopausal ET use and ovarian cancer susceptibility
loci on disease risk.

MATERIALS AND METHODS
All studies included in this analysis had approval from ethics committees and
written informed consent was obtained from all study participants.

Study populations
A total of 10 case-control studies participating in the OCAC
(http://apps.ccge.medschl.cam.ac.uk/consortia/ocac/index.html) were included in this
analysis, with seven in the United States and three in Europe. Specific details for each of
these studies have been published elsewhere (15-25), but their main study characteristics
are presented in Table 6.1.
We had a total of 5,403 serous and endometrioid cases and 13,337 controls across
the 10 OCAC studies; only serous and endometrioid cases were included as most studies
have shown that only these histotypes are significantly associated with ET use (2, 5, 26).
However, only a proportion of these women had genetic data available, leaving us with
3,855 cases and 9,593 controls. Further exclusions included the following: women who
were less than 50 years of age at reference date, which was typically the date of diagnosis
for cases and the date of interview for controls, (871 cases and 2,532 controls), had past
diagnoses of cancer (other than non-melanoma skin cancer) (398 cases and 887 controls),
140
had unknown or missing HT information (171 cases and 365 controls), or had used HT in
a combined estrogen-progestin form (664 cases and 1,758 controls). Hence, our final
dataset included 1,414 serous cases, 337 endometrioid cases, and 4,051 controls.

Genotype data
To date, 18 confirmed, genome-wide significant ovarian cancer susceptibility loci
(P≤5.0 x 10
-8
) have been identified (6-13). However, subsequent fine mapping efforts
have shown that in some instances, the originally published best “hit” in the confirmed
region was no longer the most strongly associated single nucleotide polymorphism
(SNP). Table 6.2 presents the originally published SNPs and, where applicable, the
current best hits, which we used in the analysis presented here (6).
Details regarding the genetic data have been previously described (9). Briefly,
existing genotype data from three GWASs, their replication efforts, and two large-scale
arrays (the Collaborative Oncological Gene-Environment Study (iCOGS) and the Exome
chip) were combined with data from the April 2012 release of the 1000 Genomes Project
and imputation using the program IMPUTE2 (27) was carried out for all OCAC
participants. Subjects from two studies (North Carolina Ovarian Cancer Study and New
England Case-Control Study of Ovarian Cancer) were split into two analytic sets based
on the varying scope of genotype data (genome-wide versus array) available for
imputation. This resulted in a total of 12 analytic sets for analysis (see Table 6.1
footnote).

141
Exposure and covariate data
Self-completed questionnaires and phone or in-person interviews were used to
collect information on HT use and other potential confounding variables including age,
OC use, parity, hysterectomy, tubal ligation, endometriosis, and education. Given that use
of ET increases risk of endometrial cancer in women with intact uteri (28), the majority
of ET users were hysterectomized and hence, their true age at menopause was unknown.
We therefore assumed that all women in our analysis had an age at menopause of 50,
which is the average age at menopause for women in the Western world (29).
Given the importance of menopause to ovarian cancer etiology, the effects of ET
use prior to menopause when endogenous estrogen levels are naturally high could be
inherently different from its effects after menopause (30). Therefore, we only considered
women as ET users if they used ET after age 50 for at least one year. Non-users were
women who had never used ET after age 50 (including women who only used ET before
age 50) or had only used ET after age 50 for less than one year as the effect of such short-
term use is likely to be minimal. However, a sensitivity analysis was conducted using a
true “never” user baseline group, and the results did not change. Duration of
postmenopausal ET use was assessed in the following categories: 1 to <5 years, 5 to <10
years, and 10+ years.

Statistical analysis
All models were conditioned on analytic set, five-year age category (50-54, 55-
59, 60-64, 65-69, 70-74, and 75+ years), and genetic ancestry (European, Asian, African,
and other) as determined by the Local Ancestry in Admixed Populations program (31).
142
Women with >90% European ancestry were classified as European, >80% Asian or
African ancestry were classified as Asian or African, respectively, and those with mixed
ancestry were classified as other (9). In addition, all models were adjusted for OC use
(never [including <1 year of use], 1 to <2 years, 2 to <5 years, 5 to <10 years, and 10+
years), parity (never, 1 birth, 2+ births), hysterectomy (yes/no), endometriosis (yes/no),
tubal ligation (yes/no), and education (less than high school, high school, some college,
college graduate or higher) since they were judged to be potentially important
confounders a priori. Missing categories were created for women missing any of the
covariates so their data could be included in the analysis. Data on hysterectomy status
was not available from all sites, but sensitivity analyses showed that hysterectomy status
did not substantially impact the estimates for ET or any of the SNPs.
Weighted genetic risk scores, which took into account the 18 confirmed SNPs
simultaneously, were calculated by taking the beta coefficients for each SNP’s
association with risk of serous and endometrioid ovarian cancer using all OCAC studies
in which genotype data was available (43 OCAC studies, which included 18,174 cases
and 26,134 controls (9)) and multiplying them by the genotype value (0-2) for each
subject (i.e., beta coefficients were derived from a much larger dataset). These values for
the 18 SNPs were then summed to obtain each individual’s total risk score, which was
then categorized into quartiles according to the distribution in controls for ease of
interpretation.
ORs and 95% CIs were calculated for the main effect association between each
SNP or genetic risk score quartile and disease risk using conditional logistic regression.
This was done for the serous and endometrioid histotypes separately. Previous analyses
143
that evaluated ET’s main effect on risk of serous ovarian cancer showed no difference by
grade so all serous cases were combined in our analysis (5). These associations were
further stratified by whether or not ET was used after age 50. To test for statistical
interaction (i.e., departure from a multiplicative model) between the 18 ovarian cancer
susceptibility loci or risk score and ET use on risk of serous and endometrioid ovarian
cancer, the likelihood ratio test was used to compare models with and without interaction
terms. A similar approach was used to analyze the effect of duration categories of ET use
for the associations showing a significant interaction with ever/never ET use.
All P-values reported were two-sided and considered significant at P≤0.05. An
adjusted P-value that factored in the number of tests for interaction conducted was
considered as well. All analyses were performed using SAS version 9.4 (SAS Institute,
Cary, North Carolina).

RESULTS
A total of 5,802 women were included in these analyses, with 1,414 serous cases,
337 endometrioid cases, and 4,051 controls (Table 6.1). Approximately 13.6%, 20.0%,
and 15.1% of the controls, serous cases, and endometrioid cases reported using ET after
age 50. In addition, 18 confirmed ovarian cancer SNPs were investigated here and their
characteristics are presented in Table 6.2; for 10 of the 18 SNPs, their corresponding
previously reported best hits are listed as well.
Table 6.3 shows the main effects of the 18 SNPs in the current analysis as well as
in the full OCAC dataset for the serous histotype. Their effects as well as the effects of
genetic risk score in quartiles when stratified by previous use of ET are presented in
144
Table 6.4. There was a statistically significant interaction between ET use and the T allele
of rs10069690 on chromosome 5 on risk of serous ovarian cancer at a P≤0.05 level (P-
interaction=0.013) (Table 6.4). While the T allele of rs10069690 was associated with a
48% increased risk of serous ovarian cancer among ET users (OR=1.48, 95% CI: 1.15,
1.91), there was essentially no risk among non-users (OR=1.09, 95% CI: 0.97, 1.22).
Tables 6.5 and 6.6 presents the same information as Tables 6.3 and 6.4,
respectively, but for the endometrioid histotype. Two statistically significant interactions
between the genetic variants and ET use on risk of disease were observed (Table 6.6).
Rs7207826 (T allele) on chromosome 17 was positively associated with the endometrioid
histotype among non-users of ET (OR=1.33, 95% CI: 1.09, 1.62, P-interaction=0.021),
but showed a decreased risk of disease among ET users (OR=0.53, 95% CI: 0.28, 0.94).
Similarly, non-users of ET carrying the C allele for rs56318008 on chromosome 1
showed an increased risk of endometrioid ovarian cancer (OR=1.54, 95% CI: 1.22, 1.94,
P-interaction=0.037), whereas ET users showed a decreased risk although this effect
estimate did not reach statistical significance (OR=0.74, 95% CI: 0.39, 1.41). Genetic risk
score did not appear to interact with ET use on risk of either histotype (P-interaction
=0.52 for serous, P-interaction=0.25 for endometrioid) (Tables 6.4 and 6.6).
For each of the three SNPs that showed a statistically significant interaction with
postmenopausal ET use on serous or endometrioid ovarian cancer risk at a P≤0.05 level,
the association between the SNP and risk of disease was assessed by duration of ET use.
Rs7207826 and rs56318008 did not have significant interactions with duration for
endometrioid ovarian cancer (P-interaction=0.087 and 0.18, respectively). However,
rs10069690 did have a significant interaction for serous ovarian cancer (P-
145
interaction=0.034); among women who had used ET for 10+ years, those who carried the
T allele had over a two-fold increased risk relative to those who carried the C (reference)
allele (OR=2.08, 95% CI: 1.36, 3.19) (Table 6.7).
With 18 SNPs plus a genetic risk score for two histotypes and three additional
duration interactions, we conducted a total of 41 tests for interaction in the analyses
presented here. Four of these interactions were considered statistically significant at a
P≤0.05 level. Although this is twice as many interaction associations as would be
expected by chance at the P≤0.05 level, none of the them met a Bonferroni threshold for
multiple comparisons of P=1.22 x 10
-3
(0.05/41 tests).

DISCUSSION
We have shown evidence of statistical interactions between postmenopausal ET
use and three confirmed ovarian cancer susceptibility alleles with risk of serous and
endometrioid ovarian cancer. Although none of the interactions we report here remained
significant after adjusting for multiple comparisons, the most significant interaction
identified was rs10069690 for serous ovarian cancer, a SNP whose main effect happens
to have only been observed for the serous histotype (13).
Rs10069690 is located in the TERT gene, which encodes a key component of the
enzyme telomerase known to help maintain telomere length. The expression of TERT has
been shown to be upregulated by estrogen (32), an observation that is consistent with the
results here given that the increased risk seen is among ET users. In addition, Killedar et
al reported rs10069690 as a likely functional SNP since the risk-associated T allele
results in a mRNA splice variant that encodes a catalytically inactive protein that acts as
146
an inhibitor of telomerase activity (33). The decreased enzymatic activity may result in
shorter telomeres, which could lead to an increased risk of genetic instability and
subsequent carcinogenesis (33). If greater expression of TERT from higher levels of
estrogen leads to a greater decrease in telomerase activity, then this could explain why
the association between the T allele of this SNP and serous ovarian cancer is substantially
elevated among women with longer durations of ET use (OR=2.08, 95% CI: 1.36, 3.19
for 10+ years).
The additional two interactions observed with ET use were rs56318008 and
rs7207826 for endometrioid ovarian cancer. Rs56318008 is located near WNT4, a gene
that plays a key role in steroidogenesis (34) and has been implicated in GWASs for risk
of endometriosis (35), an estrogen-related gynecologic condition strongly associated with
the endometrioid histotype (36). In addition, studies have suggested that estrogen induces
WNT4 expression (37), although the precise mechanism by which this occurs is currently
not known nor is it clear that WNT4 is the target of risk SNPs at this locus. Rs7207826 is
located near SKAP1, a gene that is primarily involved in T cell signaling and the
regulation of the lymphocyte function-associated antigen 1 gene (LFA-1). Although
expression and regulation of SKAP1 does not appear to be directly related to female sex
hormones, some studies have shown that SNPs in SKAP1 that are associated with
estrogen and androgen receptor-binding sites predict prostate cancer survival (38, 39), a
hormone-related malignancy that shares common susceptibility regions with ovarian
cancer (40). Again, it should be noted that SKAP1 has not been shown to be the target of
risk SNPs at this locus.
147
Although this study is the largest of its kind, it still has a modest sample size in
which to attempt to discover interactions. In addition, the self-reported nature of the
exposure and covariate data used could be considered a limitation. However, studies have
shown high agreement between information collected using interviews versus records for
HT use (41) as well as other reproductive factors (42, 43). Our results may be due to
chance as these interactions do not survive correction for multiple hypothesis testing, but
the fact that these are confirmed susceptibility alleles adds support to our findings. In
addition, given the apparent importance of estrogen regulation in TERT expression,
rs10069690 is of particular interest. From a biological standpoint, this SNP appears to
affect telomerase activity and hence, telomere maintenance, actions that could promote
tumorigenesis if improperly regulated (33). Although we cannot rule out that the
observed interaction may be due to a SNP in the region that is in linkage disequilibrium
with rs10069690, the fact that rs10069690 is functional with biological plausibility
supporting its interaction with ET use makes it a strong candidate. The other two SNPs
implicated in this analysis are intriguing as well in that they are confirmed ovarian cancer
susceptibility loci located in or adjacent to genes in which estrogen may be involved.
However, as previously mentioned, the target genes for these SNPs are unknown and
hence their relevance remains uncertain.
In conclusion, our results highlight the complexity of ovarian cancer etiology. The
roles that ET and the 18 ovarian cancer common variants play in ovarian carcinogenesis
may be beyond their independent effects. This is the first study, to our knowledge, to
suggest potential gene-environment interactions in ovarian cancer in the context of HT
148
use with confirmed susceptibility alleles. These findings, if replicated, may be critical for
future risk prediction modeling.

149
Table 6.1. Description of Studies Included in Analysis

Study Name Country
Study
Period
Method of Data
Collection
Number of
Controls
a

Number of
Serous
Cases
a

Number of
Endometrioid
Cases
a

Disease of the Ovary and their
Evaluation Study (23)
USA 2002-2009 In-person interview 547 (123) 218 (65) 53 (10)
German Ovarian Cancer Study
(15)
Germany 1992-1998
Self-completed
questionnaire
232 (39) 66 (11) 11 (4)
Hawaii Ovarian Cancer Study
(17)
USA 1994-2007 In-person interview 229 (28) 68 (13) 23 (6)
Hormones and Ovarian Cancer
Prediction (18)
USA 2003-2008 In-person interview 694 (68) 168 (28) 48 (7)
Malignant Ovarian Cancer
Study (24)
Denmark 1994-1999
In-person or phone
interview
363 (47) 96 (11) 17 (2)
North Carolina Ovarian Cancer
Study (19)
b

USA 1999-2008 In-person interview 401 (70) 189 (60) 47 (10)
New England Case-Control
Study of Ovarian Cancer (21)
b

USA 1999-2008 In-person interview 394 (28) 211 (26) 56 (3)
New Jersey Ovarian Cancer
Study (25)
USA 2002-2008 Phone interview 112 (6) 63 (3) 20 (0)
United Kingdom Ovarian
Cancer Population Study (16)
United
Kingdom
2006-2007 In-person interview 490 (47) 27 (1) 9 (0)
University of Southern
California, Study of Lifestyle
and Women’s Health (20, 22)
USA 1993-2008 In-person interview 589 (95) 308 (65) 53 (9)
Total: 4,051 (551) 1,414 (283) 337 (51)
a
Number in parentheses indicates the number of estrogen-only therapy users.
b
Subjects were split into two different analytic sets.
150
Table 6.2. Characteristics of the 18 SNPs Included in the Analysis and Their Previously Reported Best Hits

SNP
a

Previously
Published Best
Hit
b

Chromosome
Band
Position
Reference
Allele(s)
c

Tested
Allele
c

Tested
Allele
Frequency
d

rs58722170 (6) 1p34.3 38096421 G C 0.15
rs10069690 (13) 5p15.33 1279790 C T 0.35
chr10:21878831:D rs1243180 (9) 10p12.31 21878831 CCCTTC 0.14
rs17329882 (6) 4q26 119949960 A C 0.15
rs1879586 rs12942666 (12) 17q21.31 43567337 C G 0.08
rs56318008 (6) 1p36 22470407 C T 0.20
rs4808075 rs2363956 (7) 19p13.11 17390291 T C 0.16
chr9:136138765:D rs635634 (6) 9q34.2 136138765 CGCCCACCACTA 0.13
rs7207826 rs9303542 (9) 17q21.32 46500673 T C 0.31
rs76837345 rs11782652 (9) 8q21.13 82668818 A G 0.04
rs62274042 rs7651446 (9) 3q25.31 156435952 G A 0.01
rs635634 (6) 9q34.2 136155000 C T 0.14
rs3744763 (10) 17q12 36090885 G A 0.69
chr17:29181220:I (6) 17q11.2 29181220 T 0.13
rs6755777 rs2072590 (8) 2q31.1 177043226 T G 0.82
rs117224476 rs3814113 (11) 9q22.2 16907967 T G 0.16
rs1400482 rs10088218 (8) 8q24.21 129541931 G A 0.09
rs116133110 (6) 6q22.1 28480635 T C 0.46
Abbreviations: SNP, single nucleotide polymorphism; OCAC, Ovarian Cancer Association Consortium.
a
Chr10:21878831:D and chr17:29181220:I are listed as rs1449962376 and rs199661266, respectively, in 1000 Genomes.
b
If not specified, the previously published best hit is the same as the current best hit considered.
c
Blank cell refers to a deletion.
d
Based on 1000 Genomes for all populations. For chr9:136138765:D (rs587729126), the tested allele frequency was based on the
controls in the full OCAC dataset since the SNP is not listed in 1000 Genomes.

151

Table 6.3. Association Between the 18 SNPs and Serous Ovarian Cancer Risk

SNP
Current Analysis Full OCAC Dataset
a

OR
b
95% CI OR 95% CI
rs58722170 1.25 1.11, 1.40 1.12 1.07, 1.17
rs10069690 1.14 1.03, 1.26 1.14 1.09, 1.19
chr10:21878831:D 1.14 1.03, 1.26 1.11 1.07, 1.15
rs17329882 1.14 1.02, 1.28 1.11 1.07, 1.16
rs1879586 1.15 1.01, 1.30 1.17 1.12, 1.22
rs56318008 1.16 1.03, 1.31 1.12 1.06, 1.17
rs4808075 1.18 1.07, 1.31 1.19 1.15, 1.23
chr9:136138765:D 1.08 0.93, 1.26 1.17 1.11, 1.23
rs7207826 1.17 1.06, 1.29 1.15 1.11, 1.19
rs76837345 1.19 0.98, 1.44 1.25 1.17, 1.34
rs62274042 1.65 1.36, 2.01 1.59 1.48, 1.71
rs635634 1.14 1.01, 1.29 1.13 1.08, 1.18
rs3744763 0.89 0.81, 0.97 0.90 0.87, 0.93
chr17:29181220:I 0.89 0.80, 0.99 0.90 0.87, 0.94
rs6755777 0.98 0.89, 1.09 0.87 0.84, 0.90
rs117224476 0.73 0.64, 0.84 0.73 0.70, 0.77
rs1400482 0.80 0.69, 0.92 0.77 0.73, 0.82
rs116133110 0.86 0.78, 0.95 0.91 0.87, 0.94
Abbreviations: SNP, single nucleotide polymorphism; OCAC, Ovarian Cancer
Association Consortium; OR, odds ratio; CI, confidence interval
a
The full OCAC dataset consists of 43 studies, which includes 18,174 cases and 26,134
controls.
b
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10,
10+ years), parity (0, 1, 2+ births), tubal ligation, hysterectomy, endometriosis, and
education (less than high school, high school graduate, some college, college graduate or
more); conditioned on age (50-54, 55-59, 60-64, 65-69, 70-74, 75+), genetic ancestry
(European, African, Asian, other), and analytic set.

152

Table 6.4. Association Between the 18 SNPs and Risk Score and Serous Ovarian Cancer
Risk, Stratified by ET Use After Age 50

No ET Use
a
ET Use
a
P Value for
Interaction OR
b
95% CI OR
b
95% CI
SNP
rs58722170 1.21 1.06, 1.38 1.40 1.04, 1.89 0.31
rs10069690 1.09 0.97, 1.22 1.48 1.15, 1.91 0.013
*

chr10:21878831:D 1.15 1.03, 1.28 1.17 0.90, 1.52 0.68
rs17329882 1.16 1.02, 1.31 1.03 0.78, 1.36 0.47
rs1879586 1.17 1.02, 1.34 0.93 0.69, 1.26 0.44
rs56318008 1.23 1.07, 1.41 0.95 0.68, 1.31 0.098
rs4808075 1.20 1.07, 1.34 1.12 0.88, 1.44 0.64
chr9:136138765:D 1.14 0.96, 1.35 0.88 0.62, 1.27 0.21
rs7207826 1.20 1.07, 1.34 1.05 0.82, 1.35 0.39
rs76837345 1.27 1.03, 1.56 0.87 0.54, 1.40 0.14
rs62274042 1.57 1.27, 1.95 2.20 1.34, 3.60 0.40
rs635634 1.17 1.02, 1.34 1.00 0.74, 1.33 0.56
rs3744763 0.90 0.81, 0.99 0.86 0.69, 1.08 0.97
chr17:29181220:I 0.88 0.78, 0.99 0.96 0.74, 1.25 0.71
rs6755777 0.99 0.89, 1.10 1.00 0.79, 1.26 0.93
rs117224476 0.76 0.66, 0.89 0.62 0.44, 0.86 0.26
rs1400482 0.81 0.69, 0.95 0.72 0.50, 1.03 0.79
rs116133110 0.88 0.79, 0.98 0.79 0.62, 1.01 0.69
Risk Score Quartile
2
nd
vs. 1
st
quartile 1.18 0.94, 1.48 0.93 0.56, 1.55
0.52 3
rd
vs. 1
st
quartile 1.52 1.23, 1.88 1.65 1.00, 2.72
4
th
vs. 1
st
quartile 2.29 1.86, 2.82 2.00 1.25, 3.19
Abbreviations: SNP, single nucleotide polymorphism; ET, estrogen-only therapy; OR,
odds ratio; CI, confidence interval
*
P ≤0.05
a
Among non-ET users, there were 1,131 cases and 3,500 controls. Among ET users,
there were 283 cases and 551 controls.
b
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+
years), parity (0, 1, 2+ births), tubal ligation, hysterectomy, endometriosis, and education
(less than high school, high school graduate, some college, college graduate or more);
conditioned on age (50-54, 55-59, 60-64, 65-69, 70-74, 75+), genetic ancestry (European,
African, Asian, other), and analytic set.

153

Table 6.5. Association Between the 18 SNPs and Endometrioid Ovarian Cancer Risk

SNP
Current Analysis Full OCAC Dataset
a

OR
b
95% CI OR 95% CI
rs58722170 0.95 0.76, 1.20 0.94 0.87, 1.02
rs10069690 0.98 0.81, 1.20 0.99 0.92, 1.07
chr10:21878831:D 1.08 0.89, 1.30 1.07 1.00, 1.14
rs17329882 1.07 0.87, 1.32 1.09 1.01, 1.18
rs1879586 0.97 0.76, 1.24 1.12 1.03, 1.21
rs56318008 1.40 1.13, 1.74 1.09 1.00, 1.19
rs4808075 1.01 0.83, 1.22 0.97 0.90, 1.04
chr9:136138765:D 0.99 0.74, 1.31 1.15 1.04, 1.27
rs7207826 1.21 1.01, 1.45 1.09 1.02, 1.17
rs76837345 1.25 0.89, 1.75 1.07 0.94, 1.21
rs62274042 1.12 0.75, 1.68 1.28 1.13, 1.46
rs635634 1.04 0.83, 1.31 1.12 1.03, 1.21
rs3744763 1.06 0.89, 1.26 1.00 0.94, 1.07
chr17:29181220:I 0.81 0.66, 0.99 0.88 0.82, 0.95
rs6755777 1.01 0.84, 1.21 0.90 0.84, 0.96
rs117224476 0.79 0.62, 1.02 0.91 0.83, 0.99
rs1400482 0.98 0.76, 1.26 0.92 0.84, 1.01
rs116133110 1.04 0.87, 1.24 0.95 0.89, 1.02
Abbreviations: SNP, single nucleotide polymorphism; OCAC, Ovarian Cancer
Association Consortium; OR, odds ratio; CI, confidence interval
a
The full OCAC dataset consists of 43 studies, which includes 18,174 cases and 26,134
controls.
b
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10,
10+ years), parity (0, 1, 2+ births), tubal ligation, hysterectomy, endometriosis, and
education (less than high school, high school graduate, some college, college graduate or
more); conditioned on age (50-54, 55-59, 60-64, 65-69, 70-74, 75+), genetic ancestry
(European, African, Asian, other), and analytic set.

154

Table 6.6. Association Between the 18 SNPs and Risk Score and Endometrioid Ovarian
Cancer Risk, Stratified by ET Use After Age 50

No ET Use
a
ET Use
a
P Value for
Interaction OR
b
95% CI OR
b
95% CI
SNP
rs58722170 0.97 0.75, 1.25 0.92 0.51, 1.60 0.73
rs10069690 0.94 0.76, 1.17 1.53 0.91, 2.58 0.20
chr10:21878831:D 1.03 0.84, 1.27 1.42 0.82, 2.44 0.38
rs17329882 1.12 0.89, 1.40 0.82 0.45, 1.50 0.48
rs1879586 1.00 0.77, 1.30 0.84 0.46, 1.57 0.59
rs56318008 1.54 1.22, 1.94 0.74 0.39, 1.41 0.037
*

rs4808075 1.05 0.85, 1.29 0.85 0.50, 1.44 0.84
chr9:136138765:D 1.03 0.75, 1.41 0.66 0.30, 1.45 0.41
rs7207826 1.33 1.09, 1.62 0.51 0.28, 0.94 0.021
*

rs76837345 1.14 0.77, 1.68 1.51 0.70, 3.28 0.35
rs62274042 1.07 0.69, 1.67 1.15 0.35, 3.76 0.73
rs635634 1.09 0.85, 1.40 0.71 0.37, 1.36 0.28
rs3744763 1.07 0.89, 1.29 0.92 0.58, 1.46 0.79
chr17:29181220:I 0.80 0.64, 1.01 0.85 0.48, 1.49 0.97
rs6755777 0.98 0.80, 1.20 1.28 0.78, 2.12 0.56
rs117224476 0.85 0.65, 1.11 0.47 0.22, 0.98 0.21
rs1400482 0.91 0.69, 1.21 1.32 0.65, 2.67 0.26
rs116133110 1.07 0.88, 1.30 0.80 0.49, 1.31 0.32
Risk Score Quartile
2
nd
vs. 1
st
quartile 1.40 0.94, 2.09 1.37 0.51, 3.65
0.25 3
rd
vs. 1
st
quartile 1.65 1.12, 2.43 1.58 0.60, 4.21
4
th
vs. 1
st
quartile 1.84 1.26, 2.69 0.84 0.30, 2.33
Abbreviations: SNP, single nucleotide polymorphism; ET, estrogen-only therapy; OR,
odds ratio; CI, confidence interval
*
P ≤0.05
a
Among non-ET users, there were 286 cases and 3,500 controls. Among ET users, there
were 51 cases and 551 controls.
b
Adjusted for oral contraceptive use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+
years), parity (0, 1, 2+ births), tubal ligation, hysterectomy, endometriosis, and education
(less than high school, high school graduate, some college, college graduate or more);
conditioned on age (50-54, 55-59, 60-64, 65-69, 70-74, 75+), genetic ancestry (European,
African, Asian, other), and analytic set.
155
Table 6.7. Association Between Rs10069690 and Serous Ovarian Cancer Risk Among ET Users, Stratified by Duration of ET Use
After Age 50

SNP
1 to <5 years
a
5 to <10 years
a
10+ years
a
P Value for
Interaction OR
b
95% CI OR
b
95% CI OR
b
95% CI
rs10069690 1.08 0.59, 1.98 1.27 0.74, 2.18 2.08 1.36, 3.19 0.034
*

Abbreviations: SNP, single nucleotide polymorphism; ET, estrogen-only therapy; OR, odds ratio; CI, confidence interval
*
P ≤0.05
a
Among those who used ET for 1 to <5, 5 to <10, and 10+ years, there were 70 cases and 193 controls, 82 cases and 168 controls, and
131 cases and 190 controls, respectively.
b
The reference group for each OR presented consists of women who used ET for each specified duration category and carried the C
(reference) allele for rs10069690. All models adjusted for OC use (never (including <1), 1 to <2, 2 to <5, 5 to <10, 10+ years), parity
(0, 1, 2+ births), hysterectomy, endometriosis, tubal ligation, and education (less than high school, high school graduate, some college,
college graduate or more); conditioned on age (50-54, 55-59, 60-64, 65-69, 70-74, 75+), genetic ancestry (European, African, Asian,
other), and analytic set.

156
REFERENCES
1. Sprague BL, Trentham-Dietz A, Cronin KA. A sustained decline in
postmenopausal hormone use: results from the National Health and Nutrition
Examination Survey, 1999-2010. Obstet Gynecol. 2012;120(3):595-603.
2. Collaborative Group On Epidemiological Studies Of Ovarian C, Beral V,
Gaitskell K, et al. Menopausal hormone use and ovarian cancer risk: individual
participant meta-analysis of 52 epidemiological studies. Lancet.
2015;385(9980):1835-1842.
3. Pearce CL, Chung K, Pike MC, et al. Increased ovarian cancer risk associated
with menopausal estrogen therapy is reduced by adding a progestin. Cancer.
2009;115(3):531-539.
4. Beral V, Million Women Study C, Bull D, et al. Ovarian cancer and hormone
replacement therapy in the Million Women Study. Lancet. 2007;369(9574):1703-
1710.
5. Lee AW, Ness RB, Roman LD, et al. Association between menopausal estrogen-
only therapy and ovarian carcinoma risk. Obstet Gynecol. In press.
6. Kuchenbaecker KB, Ramus SJ, Tyrer J, et al. Identification of six new
susceptibility loci for invasive epithelial ovarian cancer. Nat Genet.
2015;47(2):164-171.
7. Bolton KL, Tyrer J, Song H, et al. Common variants at 19p13 are associated with
susceptibility to ovarian cancer. Nat Genet. 2010;42(10):880-884.
157
8. Goode EL, Chenevix-Trench G, Song H, et al. A genome-wide association study
identifies susceptibility loci for ovarian cancer at 2q31 and 8q24. Nat Genet.
2010;42(10):874-879.
9. Pharoah PD, Tsai YY, Ramus SJ, et al. GWAS meta-analysis and replication
identifies three new susceptibility loci for ovarian cancer. Nat Genet.
2013;45(4):362-370.
10. Shen H, Fridley BL, Song H, et al. Epigenetic analysis leads to identification of
HNF1B as a subtype-specific susceptibility gene for ovarian cancer. Nat
Commun. 2013;4:1628.
11. Song H, Ramus SJ, Tyrer J, et al. A genome-wide association study identifies a
new ovarian cancer susceptibility locus on 9p22.2. Nat Genet. 2009;41(9):996-
1000.
12. Permuth-Wey J, Lawrenson K, Shen HC, et al. Identification and molecular
characterization of a new ovarian cancer susceptibility locus at 17q21.31. Nat
Commun. 2013;4:1627.
13. Bojesen SE, Pooley KA, Johnatty SE, et al. Multiple independent variants at the
TERT locus are associated with telomere length and risks of breast and ovarian
cancer. Nat Genet. 2013;45(4):371-384.
14. Pearce CL, Rossing MA, Lee AW, et al. Combined and interactive effects of
environmental and GWAS-identified risk factors in ovarian cancer. Cancer
Epidemiol Biomarkers Prev. 2013;22(5):880-890.
15. Royar J, Becher H, Chang-Claude J. Low-dose oral contraceptives: protective
effect on ovarian cancer risk. Int J Cancer. 2001;95(6):370-374.
158
16. Balogun N, Gentry-Maharaj A, Wozniak EL, et al. Recruitment of newly
diagnosed ovarian cancer patients proved challenging in a multicentre biobanking
study. J Clin Epidemiol. 2011;64(5):525-530.
17. Lurie G, Terry KL, Wilkens LR, et al. Pooled analysis of the association of
PTGS2 rs5275 polymorphism and NSAID use with invasive ovarian carcinoma
risk. Cancer Causes Control. 2010;21(10):1731-1741.
18. Ness RB, Dodge RC, Edwards RP, et al. Contraception methods, beyond oral
contraceptives and tubal ligation, and risk of ovarian cancer. Ann Epidemiol.
2011;21(3):188-196.
19. Moorman PG, Calingaert B, Palmieri RT, et al. Hormonal risk factors for ovarian
cancer in premenopausal and postmenopausal women. Am J Epidemiol.
2008;167(9):1059-1069.
20. Wu AH, Pearce CL, Tseng CC, et al. Markers of inflammation and risk of ovarian
cancer in Los Angeles County. Int J Cancer. 2009;124(6):1409-1415.
21. Terry KL, De Vivo I, Titus-Ernstoff L, et al. Androgen receptor cytosine, adenine,
guanine repeats, and haplotypes in relation to ovarian cancer risk. Cancer Res.
2005;65(13):5974-5981.
22. Pike MC, Pearce CL, Peters R, et al. Hormonal factors and the risk of invasive
ovarian cancer: a population-based case-control study. Fertil Steril.
2004;82(1):186-195.
23. Bodelon C, Cushing-Haugen KL, Wicklund KG, et al. Sun exposure and risk of
epithelial ovarian cancer. Cancer Causes Control. 2012;23(12):1985-1994.
159
24. Glud E, Kjaer SK, Thomsen BL, et al. Hormone therapy and the impact of
estrogen intake on the risk of ovarian cancer. Arch Intern Med.
2004;164(20):2253-2259.
25. Bandera EV, King M, Chandran U, et al. Phytoestrogen consumption from foods
and supplements and epithelial ovarian cancer risk: a population-based case
control study. BMC Womens Health. 2011;11:40.
26. Morch LS, Lokkegaard E, Andreasen AH, et al. Hormone therapy and different
ovarian cancers: a national cohort study. Am J Epidemiol. 2012;175(12):1234-
1242.
27. Howie B, Marchini J, Stephens M. Genotype imputation with thousands of
genomes. G3 (Bethesda). 2011;1(6):457-470.
28. Grady D, Gebretsadik T, Kerlikowske K, et al. Hormone replacement therapy and
endometrial cancer risk: a meta-analysis. Obstet Gynecol. 1995;85(2):304-313.
29. Morabia A, Costanza MC. International variability in ages at menarche, first
livebirth, and menopause. World Health Organization Collaborative Study of
Neoplasia and Steroid Contraceptives. Am J Epidemiol. 1998;148(12):1195-1205.
30. Pike MC. Age-related factors in cancers of the breast, ovary, and endometrium. J
Chronic Dis. 1987;40 Suppl 2:59S-69S.
31. Sankararaman S, Sridhar S, Kimmel G, et al. Estimating local ancestry in
admixed populations. Am J Hum Genet. 2008;82(2):290-303.
32. Calado RT, Yewdell WT, Wilkerson KL, et al. Sex hormones, acting on the
TERT gene, increase telomerase activity in human primary hematopoietic cells.
Blood. 2009;114(11):2236-2243.
160
33. Killedar A, Stutz MD, Sobinoff AP, et al. A Common Cancer Risk-Associated
Allele in the hTERT Locus Encodes a Dominant Negative Inhibitor of
Telomerase. PLoS Genet. 2015;11(6):e1005286.
34. Franco HL, Dai D, Lee KY, et al. WNT4 is a key regulator of normal postnatal
uterine development and progesterone signaling during embryo implantation and
decidualization in the mouse. FASEB J. 2011;25(4):1176-1187.
35. Nyholt DR, Low SK, Anderson CA, et al. Genome-wide association meta-
analysis identifies new endometriosis risk loci. Nat Genet. 2012;44(12):1355-
1359.
36. Pearce CL, Templeman C, Rossing MA, et al. Association between endometriosis
and risk of histological subtypes of ovarian cancer: a pooled analysis of case-
control studies. Lancet Oncol. 2012;13(4):385-394.
37. Miyakoshi T, Kajiya H, Miyajima K, et al. The expression of Wnt4 is regulated
by estrogen via an estrogen receptor alpha-dependent pathway in rat pituitary
growth hormone-producing cells. Acta Histochem Cytochem. 2009;42(6):205-
213.
38. Huang CN, Huang SP, Pao JB, et al. Genetic polymorphisms in androgen
receptor-binding sites predict survival in prostate cancer patients receiving
androgen-deprivation therapy. Ann Oncol. 2012;23(3):707-713.
39. Huang CN, Huang SP, Pao JB, et al. Genetic polymorphisms in oestrogen
receptor-binding sites affect clinical outcomes in patients with prostate cancer
receiving androgen-deprivation therapy. J Intern Med. 2012;271(5):499-509.
161
40. Bahcall O. "Shared susceptibility loci for breast, prostate, and ovarian cancers".
2013. (http://www.nature.com/icogs/primer/shared-susceptibility-loci-for-breast-
prostate-and-ovarian-cancers/). (Accessed December 3, 2015).
41. Sandini L, Pentti K, Tuppurainen M, et al. Agreement of self-reported estrogen
use with prescription data: an analysis of women from the Kuopio Osteoporosis
Risk Factor and Prevention Study. Menopause. 2008;15(2):282-289.
42. Glass R, Johnson B, Vessey M. Accuracy of recall of histories of oral
contraceptive use. Br J Prev Soc Med. 1974;28(4):273-275.
43. Gentry-Maharaj A, Taylor H, Kalsi J, et al. Validity of self-reported
hysterectomy: a prospective cohort study within the UK Collaborative Trial of
Ovarian Cancer Screening (UKCTOCS). BMJ Open. 2014;4(3):e004421.

162
Chapter 7: The Effects of New Generation Hormonal Birth Control Methods on
Breast Cancer Risk Among Young Women: A Population-based Study

163
Abbreviation List:
HFI – hormone free interval
HOW – Health of Women Study
IUD – intrauterine device
LACCSP – Los Angeles County Cancer Surveillance Program
LAC+USC – Los Angeles County + University of Southern California Medical Center
NJSCR – New Jersey State Cancer Registry
OC – oral contraceptive
RDD – random digit dialing
SEER – Surveillance, Epidemiology, and End Results
U.S. – United States
USC – University of Southern California

Running Head:
Hormonal Birth Control and Breast Cancer Risk
164
Breast cancer is the most common malignancy among American women, and
there is strong evidence that hormones play a key role in its development. Well-
established risk factors, such as late age at first birth, nulliparity, early age at menarche,
and late age at menopause, are related to endogenous hormone levels. In addition, the
Women's Health Initiative trial showed a significant increased risk of breast cancer with
combined postmenopausal hormone therapy use. Epidemiologic studies have also shown
an increased risk among those currently or recently taking combined oral contraceptives
(OCs), with risk diminishing 10 years after stopping use.
Hormonal contraceptives have long been used by millions of women,
revolutionizing their reproductive life since the 1960s. However, they have significantly
changed over the years with new birth control methods and formulations, the use of non-
androgenic synthetic progestins as well as the development of OC regimens that shorten
the hormone-free interval (HFI), eliminate it (continuous), or lengthen the active pills
(extended). Given their growing popularity, it is crucial that we understand how they may
impact breast cancer risk. Few studies have assessed their long-term effects, which is
concerning given the prolonged hormonal exposure of the breast that is associated with
some of these birth control options.
To address this question, we are proposing a population-based case-control study
that targets young women between the ages of 18 and 44. The increased risk of breast
cancer has been found to be associated with only current and recent OC use, implying
that it is predominantly young women who are at risk with regards to hormonal
contraceptives. Women newly diagnosed with breast cancer will be ascertained using the
Los Angeles County Cancer Surveillance Program (LACCSP) and the New Jersey State
165
Cancer Registry (NJSCR). In addition, because case-control studies are plagued with
control selection problems that have only increased with advancing technology, we are
proposing a novel approach for control recruitment by matching each breast cancer
patient to a friend control using the social networking platform Facebook. Neighborhood
controls will be recruited as well to evaluate the feasibility of our novel approach. Birth
control use and other relevant information will be collected using a web-based
questionnaire as well as a telephone interview.
There are approximately 62 million women of reproductive age in the United
States, and over 13 million of them rely on hormonal methods for contraception.
However, birth control has significantly changed over the years. Advancements in birth
control have paved the way for new options and regimens that are growing in popularity
due to their convenience and effectiveness, but their impact on breast cancer risk is not
well-established. Our study attempts to address this important question to ensure that we
do not find ourselves unprepared to address a possible dramatic increase in breast cancer
incidence among young women. In addition, given the tremendous growth in social
networking, our proposed unique control recruitment method can potentially reshape how
we conduct epidemiologic studies in the future.

SPECIFIC AIMS
Approximately 40% of women in the United States (U.S.) who practice
contraception rely on hormonal methods, with OCs being the most popular method
among American women. Studies have shown that its current and recent use are
associated with an increased risk of breast cancer, but OCs have significantly changed
166
with new regimens being formally introduced into the OC market within the last decades.
In addition, new birth control methods, such as injectables and hormonal intrauterine
devices (IUDs), and non-androgenic progestins, such as drospirenone, have grown in
popularity as they are not only convenient and effective, but have less side effects
associated with them. Attitudes about menstruation have also changed as most women
prefer to reduce the frequency of their menstruation. Some of these new regimens
significantly prolong a woman's exposure of the breast to hormones, which could
potentially enhance her already increased risk of breast cancer.
We are proposing a population-based case-control study of women with invasive
breast cancer identified from the LACCSP and the NJSCR. Each patient will be matched
to two controls: a friend control using a novel approach with the social networking site
Facebook and a neighborhood control using a traditional systematic algorithm based on
the address of the breast cancer patient. A comprehensive web-based questionnaire,
partnered with the Health of Women Study (HOW), a program of the Dr. Susan Love
Research Foundation, and a telephone interview will be used to assess medical history,
birth control use, and other factors related to breast cancer risk.
Our specific aims are the following:
Aim 1. Recruit 1782 English-speaking women diagnosed with invasive breast cancer
between the ages of 20 and 44 who are currently residing in either Los Angeles County,
using the LACCSP, or New Jersey, using the NJSCR.
- We will obtain permission to use each patient’s Facebook friend list to recruit
an appropriate individually matched friend control
167
- We will ask each patient to complete a web-based questionnaire and a
telephone interview to collect information on their current and past birth
control use as well as other relevant information.
Aim 2. Recruit 1782 individually-matched English-speaking friend controls currently
residing in Los Angeles County or New Jersey. They will be selected using a random
algorithm from the Facebook friend lists of each participating breast cancer patient.
- We will ask each friend control to complete a web-based questionnaire and a
telephone interview to collect information on their current and past birth
control use as well as other relevant information.
Aim 3. Recruit 1782 individually-matched English-speaking neighborhood controls using
a systematic algorithm based on the address of the breast cancer patient.
- We will ask each neighborhood control to complete a web-based
questionnaire and a telephone interview to collect information on their current
and past birth control use as well as other relevant information.
Aim 4. Evaluate the association between new birth control types and breast cancer risk
using conditional logistic regression. This analysis will consider duration and timing of
use, progestin type, OC regimen, and birth control method.
Hypothesis A: Use of extended, continuous, and shorter placebo OC regimens is
associated with an increased risk of breast cancer relative to non-use.
Hypothesis B: Use of birth control shots (in the form of Depo Provera) is
associated with an increased risk of breast cancer relative to non-use.
Hypothesis C: Use of hormonal IUDs (in the form of a Mirena) is associated with
an increased risk of breast cancer relative to non-use.
168
Hypothesis D: Use of OCs containing later generation progestins with lower
androgenic activity (ie. fourth generation is associated with a lower increased risk
of breast cancer relative to use of earlier generation progestins.
Aim 5. Determine the feasibility and representativeness of Facebook friend controls as a
potential control recruitment strategy.
Hypothesis A: The distributions of breast cancer risk factors, OC use, and other
relevant information between the two sets of controls are not statistically
significantly different from one another.
Hypothesis B: The results of the association between birth control use and breast
cancer risk when using friend controls only are not statistically significantly
different from the results when using neighborhood controls only.

RESEARCH STRATEGY
Significance
Over 13 million women between the ages of 15 and 44 currently use hormonal
contraceptives; this is approximately 25% of all women in the U.S. (1). The majority of
women who practice contraception use traditional cycle OCs, but new birth control
methods and OC regimens have significantly grown in popularity. The seminal
publication by the Collaborative Group of Hormonal Factors in Breast Cancer
(Collaborative Breast Group) in 1996 showed that current and recent use of OCs was
associated with an increased risk of breast cancer, implying that it is predominantly
young women (ie. women of reproductive age) who are at risk with regard to birth
control (2).
169
Women using hormonal contraceptives in this day and age face potential new
challenges when it comes to breast cancer risk as a result of new methods and changes in
OC formulations and regimens. New generation OCs include extended pills (ie. giving
active pills for 12 out of 13 weeks before a HFI or menstruation), continuous pills (ie.
giving active pills the whole cycle with no HFI or menstruation) as well as non-
androgenic progestins (ie. fourth generation progestins such as drospirenone). In addition,
there are non-oral hormonal contraceptives, such as injectables and IUDs, that have
recently grown in popularity. However, the literature on their long-term effects is
severely lacking.
The popularity of these new OCs and birth control options is growing due to their
effectiveness, convenience, and minimal side effects. A survey conducted in the U.S. in
2004 on physician attitudes toward extended cycle OC use in adolescents found that 90%
of those surveyed were already prescribing these regimens (3). In addition, the proportion
of women who had ever used injectables increased from 4.5% in 1995 to 23% in 2006 to
2010 (1). This growing trend makes it imperative that we address this research question
or we may see a significant increase in breast cancer among young women in the coming
decade.
Although breast cancer is generally uncommon among young women, it is still
the most frequent cancer among women under the age of 40, accounting for 30% to 40%
of all incident female cancers (4). A recent study on breast cancer trends in the U.S. using
data from the U.S. Surveillance, Epidemiology, and End Results (SEER) registries
between 1976 and 2009 showed that the number of new advanced cases is increasing
among women between the ages of 25 and 39 (5). This is especially troubling since
170
young women with breast cancer have been shown to exhibit more aggressive
clinicopathologic features and have lower survival rates than older women (6). They are
also presented with a variety of unique emotional and psychosocial challenges. Given
that there is currently no effective breast cancer screening tool for women under the age
of 40, it is important that we better understand the risk factors that may contribute to
early disease risk for future prevention (5).
The etiology of breast cancer is not only hereditary, but undoubtedly hormone-
related as well. Studies show that a woman's risk of breast cancer is related to her
exposure to ovarian hormones, such as estrogen and progesterone (7). This is supported
by several well-established hormonal risk factors, such as early age at menarche, late age
at menopause, and nulliparity, which all increase a woman's duration and levels of
endogenous hormonal exposure (7). Early onset of menstruation and late onset of
menopause prolong a woman's exposure to estrogen and progesterone, which are at lower
levels and absent, respectively, when a woman is postmenopausal; pregnancy reduces a
woman's lifetime number of menstrual cycles, and hence her cumulative exposure to such
hormones. The results of the Women’s Health Initiative also indicated that exogenous use
of synthetic forms of these hormones, specifically in the form of combined hormone
therapy, increased a postmenopausal woman's risk of breast cancer (8).
Figure 7.1 presents the age-incidence curve for breast cancer, which is a plot of
the logarithm of incidence against the logarithm of age. While most non-hormone
dependent cancers have a consistent straight age-incidence curve, breast cancer exhibits a
sharp slowing around the time of menopause, which indicates a premenopausal hormonal
pattern (which is characterized by a greater production of estrogen and progesterone) that
171
is associated with higher breast cancer risk (9). These higher levels of estrogen and
progesterone have also been shown to contribute to increased cell division, which plays a
role in carcinogenesis. In addition, studies of normal breast tissue have shown that the
epithelial cells of the terminal duct lobular unit, which is where the majority of breast
cancers arise, undergo major changes during the menstrual cycle. Terminal duct lobular
unit cell proliferation increases 4-fold in the mid- to late luteal phase, which is when
estrogen and progesterone levels are highest (7).
Most epidemiologic studies have shown that use of OCs appears to confer an
increased risk of breast cancer, but that there was no overall risk 10 years or more after
stopping use. This was clearly evident in the important Collaborative Breast Group’s
analysis, which implies that OCs are a risk factor for breast cancer among women of
reproductive age. This analysis also provides the most comprehensive summary of that
data, but 89% of their cases first used OCs before 1975 (10). The Nurses' Health Study II,
which is one of the most recent studies examining this association, also found an
increased risk of breast cancer with current OC use among young women, but they
focused on formulations commonly prescribed in the 1990s (11). The importance of OCs
as a risk factor has been clearly demonstrated, but the effect of new regimens has yet to
be assessed. In addition, the effects of other hormonal birth control methods, such as
hormone-releasing IUDS, on breast cancer risk are not well-established and have shown
inconsistent results as well (12, 13). Even though many of these methods have been
around since the 1990s, they have only recently grown in popularity, highlighting the
need to better understand their effects.

172
Innovation
Given the prevalence of hormonal birth control use and the growing popularity of
new types of birth control options, an assessment of their long-term effects is needed. The
Collaborative Breast Group’s analysis has shown that recent and current use of OCs is
associated with an increased risk of breast cancer, but within the last decade, new
generation OCs have entered the OC market. The impact these OCs may have on breast
cancer risk is of serious concern since the shortened or missing HFI is associated with
prolonged hormonal exposure and hence, increased cellular activity within the breast.
However, no study to our knowledge has examined this association. While most of these
new birth control methods were not formally developed until the 2000s, their popularity
has quickly grown, with some contributing a significant portion of the OC market. With
the time it takes for new medications to become more widely-used coupled with the fact
that women need to be using these OCs for at least a period of time, we are now at a point
where a proper assessment of this is feasible. In addition, we will be assessing the effects
of other hormonal birth control options, such as injectables and IUDs. These have been
around for quite some time and yet the literature is relatively limited and inconsistent.
Our proposed study will be one of the first to take an updated, comprehensive look at the
long-term effects of hormonal contraceptives.
In addition, we are proposing a novel control recruitment strategy for this
population-based case-control study. Popular traditional methods have included
neighborhood controls and random digit dialing (RDD), but with advancing technology,
these techniques have become plagued with control selection issues. Participation rates
have declined with RDD as a result of the increasing use of cellular phones, answering
173
machines, caller identifications, and multiple lines in one household, and neighborhood
controls are labor-intensive, especially in areas with restricted access and multiple
housing units (14). The inability to successfully contact those we intended to recruit leads
to concerns regarding the representativeness of our control population. However,
population-based case-control studies will undoubtedly continue to be an important
approach in characterizing health-related associations, and hence, it is important that we
consider other potential control recruitment strategies.
We are proposing to recruit friend controls using the social networking platform
Facebook as a novel alternative that is both convenient and inexpensive. Facebook and
other social networking methods have successfully been used for participant recruitment
mainly through Internet advertising (15, 16), but no study to our knowledge has used
such a technique for recruiting controls for epidemiologic studies. Given Facebook's
widespread use among the general public, and especially in our group of interest, this is
an ideal population to test the feasibility of this method. Women between the ages of 18
and 44 are the power users of Facebook, and with an age group that is constantly mobile,
using a web-based technique may be the key to successful recruitment in future studies
(17).
We will be recruiting neighborhood controls as well, which is currently one of the
"gold standard" control recruitment methods. Comparisons will be made between our
novel approach and this traditional approach to ensure that selection bias is not an issue.
Although one of the main concerns with the use of friend controls is overmatching, we
believe that the broad base of friends on Facebook and our random selection process will
minimize these overmatching issues. We hope to demonstrate that friend controls arising
174
from Facebook are no different from neighborhood controls, which are still widely-used,
and hence, can be a viable control recruitment method. Given the innovation of this
approach, our study would be the first to use such a control recruitment strategy and
would provide evidence regarding its potential and feasibility. More importantly, our
findings could reshape how we currently conduct epidemiologic studies.
We may very well see a dramatic increase in the number of breast cancer cases
among young women in the coming years if there is a greater positive association
between these new OC formulations and breast cancer risk. One study has already found
an increase in the incidence of breast cancer with distant involvement for women
between the ages of 25 and 39 using SEER data between 1976 and 2009 (5). Without any
studies assessing these relatively new exposures, we may find ourselves unprepared to
address a significant public health issue, which makes our study of utmost importance.

Approach
Our research approach is based on our previous ovarian cancer population-based
case-control study that explored the effects of various lifestyle and reproductive factors
on disease risk. However, in this study, we are interested in assessing a novel exposure
that will predominantly apply to women in a specific age group. We plan to enroll
women with newly diagnosed invasive breast cancer who are premenopausal, between
the ages of 18 and 44, and currently living in Los Angeles County or New Jersey. This
will be accomplished using the LACCSP, which is served by the University of Southern
California (USC) Keck School of Medicine and the USC/Norris Comprehensive Cancer
Center, and the NJSCR, which are both part of the National Cancer Institute SEER
175
program. The LACCSP and NJSCR are some of the most productive cancer registries in
the world with a case-reporting completion rate of 95% as of 2009 for the LACCSP. We
have also carried out high-quality epidemiological studies using the LACCSP for
decades.

The research team. We have assembled a team of experienced and dedicated
researchers who have an extensive history of collaboration with one another.
Celeste Leigh Pearce, PhD (Principal Investigator) is an Assistant Professor in
the Department of Preventive Medicine at USC's Keck School of Medicine who has a
long-standing interest in breast and ovarian cancer, in terms of both epidemiology and
clinical translational research. She has tremendous experience in running large-scale
case-control studies as she currently plays a key role in a similarly large ovarian cancer
case-control study. Dr. Pearce has also been carrying out several smaller-scale breast and
ovarian studies. Her particular interests include better understanding the etiology of these
hormone-related cancers as well as translating her research to directly impact the lives of
women who are currently living with them. To that end, she currently has a grant that
focuses on chemoprevention through the use of specific birth control medications in
women at high-risk of breast and ovarian cancer and another grant that centers on early
detection of recurrent ovarian cancer.
Malcolm C. Pike, PhD (Principal Investigator) is a Professor in the Department
of Preventive Medicine at USC's Keck School of Medicine and an attending
epidemiologist in the Department of Epidemiology and Biostatistics at Memorial Sloan-
Kettering Cancer Center. His primary research interests are the etiology and
176
chemoprevention of breast, endometrial and ovarian cancer. Dr. Pike and Dr. Pearce have
collaborated on several smaller-scale breast and ovarian studies as well as on a large
ovarian cancer case-control study. He will play a key role in recruitment of breast cancer
patients using NJSCR.
Ann S. Hamilton, PhD (Co-Investigator) is an Associate Professor in the
Department of Preventive Medicine at USC's Keck School of Medicine. She has a strong
background in cancer epidemiology and her research focuses on the etiology of breast
and germ cell cancers and on cancer outcomes related to the breast, prostate, and others.
Dr. Hamilton will be a key LACCSP collaborator; she has extensive experience in
conducting epidemiologic studies using this source. She is currently the P.I. of several
cancer registry-based NCI-SEER funded studies and is an important source to researchers
who wish to utilize the LACCSP.
Darcy V. Spicer, MD (Co-Investigator) is an Associate Professor of Clinical
Medicine in the Department of Medicine at USC's Keck School of Medicine. He is also a
medical oncologist who specializes his practice and research in the management of breast
cancer. Dr. Spicer has vast experience in the development and conduction of various
clinical studies. He will be a key clinical colleague who will provide valuable input into
the study's design and focus in terms of ensuring that its translational potential is met.
Alice W. Lee, MPH (Graduate Student) is a doctoral student of epidemiology
in the Department of Preventive Medicine at USC's Keck School of Medicine who is
currently working with Dr. Pearce. Her dissertation focuses on better understanding the
effects of hormonal exposures on breast and ovarian cancer, specifically hormone
replacement therapy. Given her research interests, this proposal is an excellent
177
opportunity for her to supplement her research with practical experience in the
epidemiology field. Ms. Lee is currently supported by a National Institute of
Environmental Health Sciences (NIEHS) T32 training fellowship in environmental
genomics, which has provided her with additional multidisciplinary training.

Eligibility criteria. This study will include English-speaking premenopausal
women between the ages of 18 and 44 who do not have a prior history of in situ or
invasive breast cancer as well as any other invasive cancer, with the exception of non-
melanoma skin cancer. Participants must have access to the Internet and be physically
and mentally capable of completing a web-based questionnaire and a telephone interview.
They must also be currently residing in Los Angeles County or the state of New Jersey at
the time of cancer diagnosis (case reference date) or at the time initial contact is initiated
(control reference date). Eligible patients will only include those who had their invasive
breast cancer histologically confirmed.
Males, postmenopausal women as well as women younger than 18 years of age or
older than 44 years of age will be excluded from this study.

Patient recruitment and enrollment. We will be recruiting women newly
diagnosed with a primary invasive breast cancer using the LACCSP and the NJSCR,
which are the population-based cancer registries covering all residents of Los Angeles
County and the state of New Jersey, respectively. Our inclusion of only incident breast
cancer patients is to ensure we captures the effects of factors contributing to disease risk
178
instead of the effects of lifestyle modifications as a result of their diagnoses or factors of
survivorship.
Breast cancer patients will be contacted by phone to see if they are interested in
volunteering for the study. Those who are not interested will be simply asked for their
main reason for not participating during this call. If the patient expresses interest, we will
mail her information regarding the study as well as the informed consent document to be
reviewed, signed, and returned. We will follow-up to answer any questions she may have
as well as provide her with instructions for accessing the web-based questionnaire; these
instructions will also be sent via e-mail. At this time, we will also discuss the Facebook
component to our study.
If the breast cancer patient has an active Facebook account, we will obtain
permission to access her friend list to recruit an appropriate friend control using a random
algorithm. Controls will be frequency matched on age (± 3 years) and race/ethnicity and
cannot be related to their corresponding case. We will contact the control via the
messaging function that is included on each individual's Facebook account. The
eligibility criteria will be outlined in the message that we send to them as well as three
links to indicate interest and participation: 1) I fulfill the eligibility criteria and would like
to participate, 2) I fulfill the eligibility criteria but I do not want to participate, 3) I do not
fulfill the eligibility criteria. For option 1, the link will direct the participant to a webpage
that will ask for her contact information so that we may call her to explain the study and
the process. We will then mail her the informed consent to be reviewed, signed, and
returned. Similar steps will be taken as the breast cancer patients, which is outlined
above. For option 2, the link will direct them to a short survey that will ask them to
179
choose a reason for not participating. For option 3, the link will direct them to a short
survey that will ask them which criteria they did not fulfill. The random algorithm will be
continued until a Facebook friend control is recruited.
We will also be recruiting individually-matched neighborhood controls for each
breast cancer patient regardless of whether or not a friend control is obtained for each.
Although young adult women are the power users of social networking, there remains the
possibility that they do not have a Facebook account or will not allow us access to their
friend list. Neighborhood controls will not only be a good control back-up in the event
Facebook is unsuccessful, but more importantly, will allow us to compare the feasibility
and representativeness of our novel control recruitment approach to a traditional one. We
will use a systematic algorithm based on the address of the patient for recruitment, which
has been done in our previous case-control studies. Those who were at home, fulfill the
eligibility criteria, and express interest in the study will be consented and given
instructions on how to access the web-based questionnaire. Letters will be left if no one is
home, and follow-up mail, telephone, and further visits will be continued until an eligible
control agrees to participate or 150 housing units has been screened.

Data collection. All study participants will be asked to register for the Health of
Women (HOW) Study, which is an international online study of breast cancer currently
being conducted at the Dr. Susan Love Research Foundation in collaboration with City of
Hope Comprehensive Cancer Center. Additional information can be found at the
following website: https://www.healthofwomenstudy.org/. Participants will be instructed
to create an account so that they can access the web-based questionnaire, which will
180
consist of questions pertaining to their health, medical history as well as other factors
related to breast cancer. This collaboration with the HOW will provide us with the
platform for data collection while the HOW recruits more participants for their ongoing
study. After the online questionnaire has been completed, we will follow-up with each
participant and administer a telephone interview that will ask for detailed information
about her current and past birth control use. Both the questionnaire and the interview
consist of questions that have been validated by previous studies.
Data will also be gathered regarding each patient's Facebook usage prior to
administering the questionnaire and interview. This will include specific information
regarding when the patient first created a Facebook account, her general use of it as well
as other relevant information. In addition, we will obtain data from the each patient's
registry file, which will be compared to her self-reported information and used to
complete any missing information from her web-based questionnaire and telephone
interview. Extra information obtained from the file may be considered in later analyses if
applicable. In addition, the pathology report that is included in the file will be used to
verify the carcinoma and ensure that the patient is eligible for the study.
We will also be collecting information on those who choose not participate in the
study. For the contacted friend controls who are eligible, we will collect basic data on
why they chose not to participate; for those who are ineligible, we will collect data on
which criteria they did not fulfill. This will be accomplished using the various links
outlined in the Facebook message used to initiate contact. For the breast cancer patients,
we will simply ask them during our study introduction call to determine the reason for
their non-participation.
181
An experienced data analyst will then carefully go through each participant's
questionnaire/interview responses and code the data into variables for a raw dataset. This
will then be translated into the appropriate variables for a working dataset and cleaned for
our analyses. The analyst will also quantify the data collected on those who are not
participating.

Statistical methods and power. To evaluate the first hypothesis of Aim 4, the
different types of OCs will be categorized by specific regimens: 1) traditional 21-day
cycle with a HFI of 7 days, 2) shortened placebo with a HFI of 2 to 4 days, 3) extended
84-day cycle with a HFI of 7 days, and 4) continuous with no HFI. The association
between each type of regimen and breast cancer risk will be assessed using conditional
logistic regression with the baseline group consisting of never users of OC. This
association will also be re-assessed using the traditional 21-day cycle as the baseline. To
evaluate the second and third hypotheses of Aim 4, the association between the other
hormonal birth control options (ie. Depo Provera and Mirena) and breast cancer risk will
also be assessed similarly using conditional logistic regression with the baseline group
consisting of never users of OC. For all associations examined, we will also consider the
effects of duration as well as timing of use on breast cancer risk. Potential confounding
will be assessed and the appropriate factors will be adjusted for.
Because the analysis plan we have outlined only considers whether the woman
has ever used the specific regimens mentioned, we will conduct a sensitivity analysis that
will recode a woman's OC exposure according to the type she used the majority of her
exposure time. Since most women take multiple types of birth control throughout their
182
reproductive lifetime, the type she used for more than 50% of the total exposure time will
determine the category to which she will be included in the analysis. If the woman did
not take a specific birth control for more than 50% of the total exposure time, the
category will be based on the type with the highest amount of exposure.
To evaluate the four hypothesis of Aim 4, the different types of OCs will be
categorized by the OC generation/type of progestin used: 1) first generation (no
progestin), 2) second generation (contains levonogestrel, norgestimate, and other
members of the norethindrone family), 3) third generation (contains desogestrel or
gestodene), 4) fourth generation (contains drospirenone, dienogest, or nomegestrol
acetate). The association between the fourth generation OCs and breast cancer risk will
be assessed using conditional logistic regression with the baseline group consisting of the
earlier generation OCs. Duration and timing of use will also be considered. In addition,
because this analysis only considers whether or not the woman has ever used each
generation of OCs, we will conduct a sensitivity analysis and recode a woman's OC
exposure according to the type she used the majority of her exposure time in a similar
fashion as mentioned above.
The previously outlined analyses consider all recruited participants, including
both Facebook friend controls and neighborhood controls. Therefore, to address Aim 5,
the analyses will be re-conducted by type of control and the results will be compared.
This will allow us to determine whether our novel control recruitment strategy is a viable
alternative to the traditional neighborhood control approach. In addition, we will compare
the distributions of OC use, demographics, breast cancer risk factors, and other relevant
183
information between the friend controls and the neighborhood controls as a way to assess
the representativeness of our novel approach.
In these analyses, we will also stratify by when the Facebook account was
created. Because some patients may create a social networking account to link with
others who have similar experiences or to join support groups, this will ensure that the
friends we are randomly selecting our controls from reflect the general population and
that selection bias is not an issue in our study.
Our target enrollment of 1782 case-control pairs affords us 80% power to detect
at least a 40% increased risk of breast cancer with a two-sided alpha of 0.05. This was
determined using Quanto, with the assumption that approximately 9.0% of women of
reproductive age (ages 15-44) in the United States use new generation OCs as their
method of contraception and that a woman's lifetime risk of developing breast cancer
under the age of 40 is 0.6%. The proposed sample size also includes a 20% inflation to
account for non-respondents or missing information. We chose a 40% increased risk as
the minimum effect estimate to detect since one of the most recent studies that looked at
OC use and breast cancer risk among young women found an effect estimate of 1.33
among current and recent users, and we hypothesize that the effect of new generation
OCs on breast cancer risk should be greater than the traditional regimens that were
popular in the 1990s and earlier. Other epidemiologic studies have also shown similar
effect estimates as well. We present Table 7.1, which shows the number of case-control
pairs needed to detect each tenth of an increase in the risk estimate.

184
Timeline. This project will take approximately three years to complete. The first
two years will be devoted to recruitment of invasive breast cancer patients and
individually-matched controls. In 2010, there were 1443 invasive breast cancer diagnoses
among women between the ages of 20 and 44 in Los Angeles County and New Jersey,
according to SEER*Stat Software Version 8.1.2 (18). During this time, we will also be
collecting information on their OC use and other relevant information from our web-
based questionnaire and telephone interview. The last year will be devoted to analyzing
the data.

185
Table 7.1. Sample Size Needed by Risk Estimate With a Power=80% and a Two-sided
Alpha=0.05

Risk Estimate
# of Case-
Control Pairs
20% Inflation
Final # of Case-
Control Pairs
1.40 1485 297 1782
1.50 997 200 1197
1.60 725 145 870
1.70 557 112 669

186
Figure 7.1. Age-incidence Curve of Breast Cancer Using Data from White Women in the
United States, 1969-1971 (9)

Logarithm of incidence (y-axis) versus logarithm of age (x-axis).

187
REFERENCES
1. Guttmacher Institute. " Contraceptive Use in the United States. " 2015.
(https://www.guttmacher.org/fact-sheet/contraceptive-use-united-states).
(Accessed January 5, 2016).
2. Collaborative Group on Hormonal Factors in Breast C. Breast cancer and
hormonal contraceptives: collaborative reanalysis of individual data on 53 297
women with breast cancer and 100 239 women without breast cancer from 54
epidemiological studies. Lancet. 1996;347(9017):1713-1727.
3. Gerschultz KL, Sucato GS, Hennon TR, et al. Extended cycling of combined
hormonal contraceptives in adolescents: physician views and prescribing
practices. J Adolesc Health. 2007;40(2):151-157.
4. National Cancer Institute. "Cancer Epidemiology in Older Adolescents and
Young Adults 15 to 29 Years of Age, Including SEER Incidence and Survival:
1975-2000." 2006.
(http://www.seer.cancer/gov/archive/publications/aya/aya_mono_complete/pdf).
(Accessed July 20, 2014).
5. Johnson RH, Chien FL, Bleyer A. Incidence of breast cancer with distant
involvement among women in the United States, 1976 to 2009. JAMA.
2013;309(8):800-805.
6. Fredholm H, Eaker S, Frisell J, et al. Breast cancer in young women: poor
survival despite intensive treatment. PLoS One. 2009;4(11):e7695.
7. Spicer DV, Pike MC. The prevention of breast cancer through reduced ovarian
steroid exposure. Acta Oncol. 1992;31(2):167-174.
188
8. Rossouw JE, Anderson GL, Prentice RL, et al. Risks and benefits of estrogen plus
progestin in healthy postmenopausal women: principal results From the Women's
Health Initiative randomized controlled trial. JAMA. 2002;288(3):321-333.
9. Pike MC, Krailo MD, Henderson BE, et al. 'Hormonal' risk factors, 'breast tissue
age' and the age-incidence of breast cancer. Nature. 1983;303(5920):767-770.
10. Breast cancer and hormonal contraceptives: collaborative reanalysis of individual
data on 53 297 women with breast cancer and 100 239 women without breast
cancer from 54 epidemiological studies. Collaborative Group on Hormonal
Factors in Breast Cancer. Lancet. 1996;347(9017):1713-1727.
11. Hunter DJ, Colditz GA, Hankinson SE, et al. Oral contraceptive use and breast
cancer: a prospective study of young women. Cancer Epidemiol Biomarkers Prev.
2010;19(10):2496-2502.
12. Dinger J, Bardenheuer K, Minh TD. Levonorgestrel-releasing and copper
intrauterine devices and the risk of breast cancer. Contraception. 2011;83(3):211-
217.
13. Lyytinen HK, Dyba T, Ylikorkala O, et al. A case-control study on hormone
therapy as a risk factor for breast cancer in Finland: Intrauterine system carries a
risk as well. Int J Cancer. 2010;126(2):483-489.
14. Bernstein L. Control recruitment in population-based case-control studies.
Epidemiology. 2006;17(3):255-257.
15. Fenner Y, Garland SM, Moore EE, et al. Web-based recruiting for health research
using a social networking site: an exploratory study. J Med Internet Res.
2012;14(1):e20.
189
16. Kapp JM, Peters C, Oliver DP. Research recruitment using Facebook advertising:
big potential, big challenges. J Cancer Educ. 2013;28(1):134-137.
17. Smith J. "December Data on Facebook’s US Growth by Age and Gender: Beyond
100 Million". 2010. (http://www.insidefacebook.com/2010/01/04/december-data-
on-facebook%E2%80%99s-us-growth-by-age-and-gender-beyond-100-million/).
(Accessed August 1, 2014).
18. Surveillance, Epidemiology, and End Results (SEER) Program Research Data
(1973-2012) Version 8.2.1. National Cancer Institute, DCCPS, Surveillance
Research Program, Surveillance Systems Branch; 2012.
(http://seer.cancer.gov/seerstat/software/) (Accessed February 2, 2016).

190
Chapter 8: Conclusion

191
Abbreviation List:
EPT – estrogen-progestin combined therapy
ET – estrogen-only therapy
HT – hormone therapy
OC – oral contraceptive
OCAC – Ovarian Cancer Association Consortium

Running Head:
Conclusion

192
The goal of this dissertation was to explore the effects of hormonal exposures on
ovarian and breast cancer risk, two malignancies whose etiologies are influenced by
estrogen and progesterone. Endogenous levels of these hormones significantly change
throughout a woman’s lifetime. However, hormone therapy (HT) and oral contraceptives
(OCs), which consist of synthetic forms of estrogen and progesterone, alter what the
female body experiences naturally for birth control before menopause and menopausal
symptom relief and osteoporosis prevention after menopause. While use of such
exogenous hormones improves a woman’s quality of life, their effects on future risk of
hormone-related malignancies are important to understand given that ovarian cancer is
the most fatal female gynecologic cancer and breast cancer is the most common female
cancer.
There were two key research components to this dissertation. First, the analyses
presented a comprehensive evaluation of the HT-ovarian cancer association with
considerations for HT type, timing and duration of HT use, and disease histotype. An
assessment of potential gene-environment interactions in the context of estrogen-only
therapy (ET) and confirmed ovarian cancer susceptibility loci was included as well. This
used multi-site primary data from the Ovarian Cancer Association Consortium (OCAC),
which allowed for a better evaluation of both unaddressed HT questions as well as
inconsistent HT findings across previously published studies. Second, the grant proposed
an epidemiologic study that not only would add to what we currently know about the OC-
breast cancer association, but would explore how recent changes in hormonal
contraception might affect it. With the growing popularity of new OC formulations that
minimize monthly menstruations, their impact on breast cancer risk is an important
193
research question, especially given the prolonged hormonal exposure of the breast
associated with their use. In addition, most of these new generation birth control methods
only formally entered the contraception market in the last decade or two, making a new
epidemiologic study the optimal way to address such a question.

SUMMARY OF CHAPTERS 4-6
Chapter 4
The fourth chapter presented a detailed evaluation of the association between use
of ET and risk of ovarian cancer using questionnaire data from 10 OCAC case-control
studies. The results were based on analyses restricted to hysterectomized women only
given that ET is a confirmed risk factor for endometrial cancer and hence is
contraindicated for women with intact uteri (1). Focusing on these women was thought to
be appropriate since the findings would be most applicable to the population most likely
to use ET. These analyses also focused on exclusive ET use in an effort to best estimate
ET’s independent effects. In addition, unlike most previously published studies, only ET
use after the age of 50 was considered. From a biological standpoint, the added estrogen
exposure on risk of ovarian cancer may be more relevant during a period when
endogenous estrogen levels are naturally lower (ie. after menopause); this may be
especially true given that menopause plays an important role in ovarian cancer etiology
(2). Because women who have had a hysterectomy will not be able to determine their age
at menopause due to the cessation of their monthly menstruations, age 50, the average
age at menopause for women in these populations (3), was used.
194
The results indicated that not only was there a significant increased risk of ovarian
cancer with postmenopausal ET use, but that this increased risk was characterized by
distinct recency and duration of use effects as well as histotype-specific effects. The
greatest relative risk observed was over fourfold among current ET users of 10+ years for
the endometrioid histotype. Significant increased risks were also seen for the serous
histotype, but not for the mucinous and clear cell histotypes. However, what was most
striking was the duration effect as the results for overall, serous, and endometrioid
ovarian cancer showed clear risk differences between short-term (ie. 1 to <5 years) and
long-term (ie. 10+ years) ET use. In fact, short-term ET use appeared to have little risk
associated with it whereas long-term ET use not only showed a significant elevation in
risk, but an increased risk that persisted even after ET was last used more than five years
prior to a woman’s date of diagnosis or interview.
What Chapter 4 highlighted was the unequivocal nature of ET’s association with
risk of disease. In addition, it confirmed ET’s histotype-specific effects with serous and
endometrioid ovarian cancer, which were reported in previous studies including the
pooled analysis by Collaborative Group on Epidemiological Studies of Ovarian Cancer
(Collaborative Group), the most comprehensive assessment of the HT-ovarian cancer
association to date (4). However, the results presented in Chapter 4 showed a clear
duration effect, a finding that the Collaborative Group did not observe for overall HT use
despite the duration effects seen for ET in the primary publications that constituted the
majority of their statistical information (5,6). Overall, Chapter 4 was an attempt to
evaluate the association between ET and ovarian cancer in the most analytically sound
way so its findings would be relevant to the most appropriate population.
195
Chapter 5
While Chapter 4 focused on the effects of ET, Chapter 5 specifically examined
the independent effects of estrogen-progestin combined therapy (EPT) on risk of ovarian
cancer with similar considerations for recency and duration of use and disease histotype.
A total of 11 OCAC case-control studies were included in these analyses; all studies
overlapped with those included in Chapter 4 except for one due to missing age at
menopause information. A similar analytic approach was used to assess the EPT-ovarian
cancer association, including the exclusion of women who had used ET. However, in this
analysis, because the effect estimates when describing the EPT-ovarian cancer
association for premenopausal versus postmenopausal use were similar, EPT use was
considered regardless of when it was used relative to menopause. In addition, both
women with and without intact uteri were included as there is no contraindication that
would preclude certain women from using EPT. Subset analyses on seven of the 11
studies were also conducted to evaluate how specific EPT regimens may differentially
impact ovarian cancer risk given that the typical scheduling of concurrent progestin use
with ET may be 5 to 15 days per month (sequential EPT) or daily (continuous EPT).
Overall, use of EPT was associated with a slight decreased risk of ovarian cancer,
although this appeared to be driven by histotype-specific effects for mucinous and clear
cell tumors as well as women who reported using EPT daily. Interestingly, current long-
term EPT users of 10+ years showed an increased risk of ovarian cancer whereas past
long-term users of 10+ years showed a decreased risk, prompting some interesting
thoughts regarding progesterone’s role in ovarian cancer etiology and how it may interact
with estrogen’s risk-inducing effects. However, what was evident from this analysis was
196
the need to consider both exposure and disease characteristics when assessing EPT’s
impact on risk of ovarian cancer.
The underlying goal of Chapter 5 was to examine how ET’s association with risk
of ovarian cancer is impacted by the addition of a progestin. Progesterone is hypothesized
to play a protective role in ovarian carcinogenesis (7), hence part of the motivation for
this analysis was to determine if that hypothesis would reflect in EPT’s association with
ovarian cancer since ET was found to be associated with an increased risk. While the
results observed here were in line with what is expected biologically, the Collaborative
Group’s analysis did not see this attenuation in risk and reported comparable increased
risks for both HT types. However, what was most peculiar about the Collaborative
Group’s findings was how the increased risk associated with EPT use was only reported
for prospective studies and not retrospective studies. Overall, Chapter 5 added to the
current EPT-ovarian cancer literature while highlighting the complexity of estrogen and
progesterone’s role in ovarian carcinogenesis.

Chapter 6
Given the clear association between ET use and risk of ovarian cancer, Chapter 6
further examined this by considering ET’s potential interactive effects with 18 confirmed
ovarian cancer susceptibility loci previously identified using genome-wide association
studies. Because each of these variants was found to be associated with a relatively
modest effect estimate, interactions were thought to be possible which could put some
women at higher risk, a research question that has yet to be addressed according to the
current literature. This evaluation of gene-environment interactions required both
197
epidemiologic and genetic data from 10 OCAC case-control studies; all studies
overlapped with those included in Chapter 4 except for one which did not have genotype
data on its subjects. To identify statistical interactions, the associations between each
confirmed variant and serous and endometrioid ovarian cancer were assessed according
to prior use of ET; only serous and endometrioid ovarian cancer were evaluated since the
main effects analysis as presented in Chapter 4 showed only these two histotypes to be
associated with ET.
The most intriguing finding in Chapter 6 was the modifying effect of ET use on
the association between rs10069690 and serous ovarian cancer. Among carriers of the T
allele, those who had used ET were at a significant increased risk of disease whereas non-
users showed no risk at all. Functional analyses had shown rs10069690 to be a splicing
variant located in the TERT (8), a gene necessary for telomere maintenance and whose
expression has been shown to be upregulated by estrogen (9), making its interaction with
ET use on risk of disease biologically plausible. Other significant interactions were noted
with two other confirmed ovarian cancer variants, although their functional relevance
remains uncertain at this time.
Disease risk is explained by not only genetic and non-genetic factors, but also
their interplay, making gene-environment interactions key in understanding disease
etiology. What Chapter 6 presented was evidence of such potential interactions, an
observation that may be important for lifetime risk estimation. While the identified
interactions do not survive correction for multiple comparisons, they may provide clues
as to the mechanism of action of these loci. Given that no other study to date has looked
198
at these confirmed variants in the context of HT use, Chapter 6 offered intriguing insight
into these etiologic factors that may have clinical utility in the future.

SUMMARY OF CHAPTER 7
Approximately one in eight women will develop breast cancer in their lifetime,
making it the most common female malignancy not only among all women in the United
States, but also among young women under the age of 40 (10). Younger breast cancer
patients have been shown to face lower survival rates and more aggressive
clinicopathologic features in addition to unique emotional and psychosocial challenges
(11), which make a better understanding of breast cancer risk factors among this
population important. Current and recent use of OCs are known to increase a woman’s
risk of breast cancer (12), implying that it is women of reproductive age who are at risk
of disease when it comes to birth control use.
However, newer types of contraception and OC regimens have grown in
popularity in recent years due to their convenience and effectiveness with little known as
to its long-term effects. Given that the majority of studies to date have focused on the
effects of traditional contraceptive methods, Chapter 7 outlined a new population-based
case-control study to address this important research question. While breast cancer
patients would be identified using the traditional method of cancer registries, controls
would be recruited using the social media platform Facebook, a novel method that could
reshape how epidemiologic studies are conducted in the future given the growing
challenges of control recruitment.
199
The basis to Chapter 7 was twofold. First, it proposed a new study to tackle a
research question that we are only now able to properly address given that newer
generation hormonal contraceptives only formally entered the market within the last few
decades. Second, it discussed the growing difficulties with control recruitment for case-
control studies due to advancements in technology, proposing a possible, novel
alternative if successful. The goal would be to submit this grant in the coming year as the
results that stem from this study could have an important clinical impact.

CONCLUDING REMARKS
The analyses and grant included in this dissertation focused on the effects of two
frequently used exposures on two important hormonal malignancies. Chapters 4 and 5
showed that ET is clearly associated with an increased risk of ovarian cancer, but
concurrent progestin use in the form of EPT may attenuate some of this risk; Chapter 6
indicated potential interactions among confirmed ovarian cancer variants, ET use, and
disease risk; Chapter 7 highlighted the need to further explore the positive association
between OCs and breast cancer given the growing popularity of non-traditional birth
control options. This dissertation not only sheds light on ovarian cancer etiology, but has
important clinical implications as well.
What is currently accepted regarding the HT-ovarian cancer association is based
on the Collaborative Group’s pooled analysis. However, some of the results presented in
this dissertation are contrary to their findings, which may be important from a public
health standpoint. As disease prevention moves towards lifetime risk estimation, the
distinct effects of ET and EPT use according to duration, recency, and disease histotype
200
as well as the potential interactive effects with GWAS-identified ovarian cancer variants,
as identified here, will become more and more clinically relevant. Current risk prediction
models and lifetime risk analyses have not considered such features (13,14), but as the
literature continues to grow, these models will become more comprehensive and
individualized, eventually distinguishing lifetime disease risk differences for ET versus
EPT use and short-term versus long-term use.
In order to improve prevention of both ovarian and breast cancer, the current
literature needs to be expanded, especially with regard to exposures that are not only
commonly used, but can be easily modified. This dissertation was an effort to fill in the
gaps in the literature in order to address questions that will not only contribute to our
understanding of both malignancies, but will likely have a significant clinical impact.

201
REFERENCES
1. Grady D, Gebretsadik T, Kerlikowske K, Ernster V, Petitti D. Hormone
replacement therapy and endometrial cancer risk: a meta-analysis. Obstet
Gynecol. 1995;85(2):304-313.
2. Pike MC. Age-related factors in cancers of the breast, ovary, and endometrium. J
Chronic Dis. 1987;40 Suppl 2:59S-69S.
3. Morabia A, Costanza MC. International variability in ages at menarche, first
livebirth, and menopause. World Health Organization Collaborative Study of
Neoplasia and Steroid Contraceptives. Am J Epidemiol. 1998;148(12):1195-1205.
4. Collaborative Group On Epidemiological Studies Of Ovarian C, Beral V,
Gaitskell K, et al. Menopausal hormone use and ovarian cancer risk: individual
participant meta-analysis of 52 epidemiological studies. Lancet.
2015;385(9980):1835-1842.
5. Morch LS, Lokkegaard E, Andreasen AH, Kruger-Kjaer S, Lidegaard O.
Hormone therapy and ovarian cancer. JAMA. 2009;302(3):298-305.
6. Beral V, Million Women Study C, Bull D, Green J, Reeves G. Ovarian cancer and
hormone replacement therapy in the Million Women Study. Lancet.
2007;369(9574):1703-1710.
7. Risch HA. Hormonal etiology of epithelial ovarian cancer, with a hypothesis
concerning the role of androgens and progesterone. J Natl Cancer Inst.
1998;90(23):1774-1786.
202
8. Killedar A, Stutz MD, Sobinoff AP, et al. A Common Cancer Risk-Associated
Allele in the hTERT Locus Encodes a Dominant Negative Inhibitor of
Telomerase. PLoS Genet. 2015;11(6):e1005286.
9. Calado RT, Yewdell WT, Wilkerson KL, et al. Sex hormones, acting on the
TERT gene, increase telomerase activity in human primary hematopoietic cells.
Blood. 2009;114(11):2236-2243.
10. National Cancer Institute. "Cancer Epidemiology in Older Adolescents and
Young Adults 15 to 29 Years of Age, Including SEER Incidence and Survival:
1975-2000." 2006.
(http://www.seer.cancer/gov/archive/publications/aya/aya_mono_complete/pdf).
(Accessed February 10, 2016).
11. Fredholm H, Eaker S, Frisell J, Holmberg L, Fredriksson I, Lindman H. Breast
cancer in young women: poor survival despite intensive treatment. PLoS One.
2009;4(11):e7695.
12. Collaborative Group on Hormonal Factors in Breast C. Breast cancer and
hormonal contraceptives: collaborative reanalysis of individual data on 53 297
women with breast cancer and 100 239 women without breast cancer from 54
epidemiological studies. Lancet. 1996;347(9017):1713-1727.
13. Li K, Husing A, Fortner RT, et al. An epidemiologic risk prediction model for
ovarian cancer in Europe: the EPIC study. Br J Cancer. 2015;112(7):1257-1265.
14. Pearce CL, Stram DO, Ness RB, et al. Population distribution of lifetime risk of
ovarian cancer in the United States. Cancer Epidemiol Biomarkers Prev.
2015;24(4):671-676.
203

Abstract (if available)

Abstract Ovarian cancer accounts for more deaths than any other cancer in the female reproductive system. With a poor survival rate due to a lack of a routine screening method, elucidating its etiology is imperative for disease prevention. The current literature suggests ovarian cancer's etiology to be hormone-related as exemplified by some of its well-established etiologic factors, such as oral contraceptive use and parity, which affect endogenous levels of estrogen and progesterone. Hormone therapy use also affects these hormone levels, but its impact on ovarian cancer risk remains to be clarified. ❧ This dissertation comprehensively examines the hormone therapy-ovarian cancer association, considering the type used, the duration and recency of its use, and disease histotype. Given that estrogen and progesterone are hypothesized to have distinct effects on the ovary and hormone therapy entails synthetic forms of these hormones, a better understanding of this association can have significant clinical implications and contribute to our understanding of ovarian carcinogenesis. This dissertation also explores potential gene-environment interactions in the context of hormone therapy, which may be useful for future lifetime risk estimation. ❧ Lastly, oral contraceptive use has been shown to increase risk of breast cancer, but how this increased risk is impacted by new birth control methods and regimens remains uncertain. This dissertation concludes with an epidemiologic study proposal to address this research question in order to evaluate the effects of another exogenous hormonal exposure on a malignancy that is often studied with ovarian cancer.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The effects of tobacco exposure on hormone levels and breast cancer risk among young women

PDF

The role of pesticide exposure in breast cancer

PDF

Prostate cancer: genetic susceptibility and lifestyle risk factors

PDF

Genes and hormonal factors involved in the development or recurrence of breast cancer

PDF

The role of heritability and genetic variation in cancer and cancer survival

PDF

Lifetime physical activity and its effects on breast cancer survival

PDF

Screening and association testing of coding variation in steroid hormone coactivator and corepressor genes in relationship with breast cancer risk in multiple populations

PDF

Diet quality and pancreatic cancer incidence in the multiethnic cohort

PDF

Factors that influence mammographic density: role of estrogen metabolism genes, biomarkers of inflammation, and lifestyle

PDF

Pathogenic variants in cancer predisposition genes and risk of non-breast multiple primary cancers in breast cancer patients

PDF

Carcinogenic exposures in racial/ethnic groups

PDF

Arm lymphedema in a multi-ethnic cohort of female breast cancer survivors

PDF

An analysis of disease-free survival and overall survival in inflammatory breast cancer

PDF

Cell non-autonomous control of proliferation in tissues of elevated cancer risk in BRCA1 mutation carriers

PDF

Associations between isoflavone soy protein (ISP) supplementation and breast cancer in postmenopausal women in the Women’s Isoflavone Soy Health (WISH) clinical trial

PDF

Instability of heart rate and rating of perceived exertion during high-intensity interval training in breast cancer patients undergoing anthracycline chemotherapy

PDF

Analysis of factors associated with breast cancer using machine learning techniques

PDF

Identifying genetic, environmental, and lifestyle determinants of ethnic variation in risk of pancreatic cancer

PDF

Utility of polygenic risk score with biomarkers and lifestyle factors in the multiethnic cohort study

PDF

Risk factors of pelvic floor disorders in the multiethnic cohort study

Asset Metadata

Creator Lee, Alice Wen-Ron (author)

Core Title The effects of hormonal exposures on ovarian and breast cancer risk

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Publication Date 04/21/2016

Defense Date 02/18/2016

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag breast cancer,Hormones,OAI-PMH Harvest,ovarian cancer

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Pearce, Celeste L. (committee chair), Roman, Lynda D. (committee member), Stram, Daniel O. (committee member), Wu, Anna H. (committee member)

Creator Email alicewle@usc.edu,alicewlee22@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-237904

Unique identifier UC11278312

Identifier etd-LeeAliceWe-4332.pdf (filename),usctheses-c40-237904 (legacy record id)

Legacy Identifier etd-LeeAliceWe-4332.pdf

Dmrecord 237904

Document Type Dissertation

Format application/pdf (imt)

Rights Lee, Alice Wen-Ron

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA