University of Southern California Dissertations and Theses

Predictive factors of breast cancer survival: a population-based study

(USC Thesis Other)

Predictive factors of breast cancer survival: a population-based study

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
Predictive factors of breast cancer survival: a population-based study

by

Jing Yu

A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOSTATISTICS AND EPIDEMIOLOGY)

May 2018

Copyright 2018 Jing Yu
ii
TABLE OF CONTENTS
DEDICATION iii
ACKNOWLEDGMENTS iv
ABSTRACT v
INTRODUCTION 1
METHODS 4
RESULTS 7
DISCUSSION 16
REFERENCES 24
TABLES 32
FIGURES 44

iii

DEDICATION
This work is dedicated to both my lovely daughter and myself.
This is something that I hope my daughter will be proud of.
This is also a precious memory of my 5-year life being as an international student at the
United States. It is an evidence of me chasing my dream.
“Now this is not the end. It is not even the beginning of the end. But it is, perhaps, the end
of the beginning.” – Churchill

iv
ACKNOWLEDGMENTS

I am very grateful to God, the Almighty King, the Creator, the Saviour and the Guardian.
Without His blessing, I couldn’t make this far.
I would like to thank my dearest parents, my husband, and my parents-in law. They have
sacrificed a lot just to let me chase my dream. They are selfless, caring, supportive and very
helpful.
I also would like to thank both of my advisors Dr. Myles Cockburn and Dr. Roberta
Mckean-Cowdin. They are both incredibly supportive and extremely helpful and
knowledgeable. I respect them very much and they are both my role models in my future
life.
I also would like to thank my committee member Dr. Meredith Franklin for her expertise
and guidance.
Finally, I would like to thank Department of Preventive Medicine at University of Southern
California for offering me a wonderful opportunity to complete my education at the United
States.

v
Abstract
Background: Breast cancer is one of the most frequently diagnosed cancers among
American women, and is one of the leading causes of death in this population. Prior
investigations suggested that breast cancer survival may be associate with age,
socioeconomic status (SES), race, stage, hormone receptor status including estrogen
receptors (ERs) and progesterone receptor (PR) and oncogene expression- human
epidermal growth factor receptor 2 (HER-2). However, most of the studies done on race
have focused on African American and White women. But the minorities of Middle
Eastern and Asian/Pacific Island descent have not been studied extensively. Therefore, with
a large, population-based study consisting of several different race/ethnic groups, we can
examine these previously under-studied characteristics. Furthermore, with a large
population we can build stronger evidence from existing information by assessing the
effects of age at diagnosis, SES, race, stage at diagnosis, ER and PR on predicting breast
cancer survival.
Methods: There were 36,740 female breast cancer cases collected from the population-
based cancer registry for Los Angeles County from 2008 to 2012. The distribution of
patients and their tumor characteristics were evaluated by age, stage and race. The Kaplan–
Meier method and the log-rank test were used to test differences in survival these sub-
groups. Cox proportional hazards regression was used to estimate hazard ratio (HR) and
its corresponding 95% confidence intervals (CI) for every independent variable
univariately, and in combination by fitting multivariate models.
vi
Results: Breast cancer survival differed by age, race, stage, ER and PR group (Log rank
p<0.0001). The survival of women who were 45-49 years old were the best compared to
other age categories. Among people between 40-70 years old, as age increased, breast
cancer survival decreased. SES was highly associated with overall breast cancer survival
with or without adjusting for other covariates. ER/PR negative patients had poorer survival
compared to ER/PR positive subjects. The survival of Middle-eastern breast cancer patients
was similar to Asian/pacific islander, and the two groups had the better survival compared
to other race/ethnicities. African American women had the worst survival.
Conclusion: We found large racial disparities in breast cancer mortality with poorest
survival for Black women the best survival for Asian/Pacific Islanders and Middle-eastern.
We also built more evidence on predictors of breast cancer survival by demonstrating 1)
women earlier breast tumor stage had better survival outcome;2) as SES level increased,
the chance of dying for the female breast cancer patients decreased. 3) Chance of dying
from breast cancer increased as age increased for women who were older than 40 years old
4) ER/PR negative also are risk factors of breast cancer mortality.

1
Introduction
Breast cancer is one of the most frequently diagnosed cancers among American women,
and is one of the leading causes of death in this population [1, 2]. About one in eight women
(12.4%) will develop breast cancer throughout the course of her lifetime in the United
States [1, 2], and about and about 21.2 per 100,000 women per year will die from breast
cancer, according to the U.S. National Cancer Institute’s Surveillance Epidemiology and
End Results (SEER) Program [1, 2]. SEER data has estimated that there were 252,710 new
cases of female breast cancer and 40,610 people died of this disease during 2017 [2].Breast
cancer has a major impact on the U.S. healthcare system due to its high incidence and
mortality rate and tremendous economic burden created by the high cost of both treatment
and the care following cancer diagnosis.
With the intention of developing prevention strategies that will both mitigate the high cost
of breast cancer treatment and improve survival outcomes, numerous research have been
conducted to investigate the determinants for breast cancer survival which includes several
main elements as followed-demographic factors, hormonal features, life-styles factors,
tumor characteristics and genetic markers.
Epidemiologic evidence suggests that age at diagnosis is associated with female breast
cancer survival- younger breast cancer patients tend to have worse survivorship compared
to elderly patients since younger patients incline to be diagnosed with more aggressive
breast cancer [3-11]. In addition, socioeconomic status (SES) is recognized to be
associated with breast cancer survival by possibly being a marker for access to care or
2
health insurance coverage [12-16]. It has been shown that SES is a potential risk factor in
terms of breast cancer survival-lower SES are less likely to have health insurance in order
to get up-to-date cancer treatment thus resulting in poorer breast cancer prognosis [12, 16,
17]. Furthermore, race/ethnicity is demonstrated to be highly associated with breast cancer
survival [16, 18-21]. Notably, African-American females have a significantly higher breast
cancer mortality rate compared to non-Hispanic white women [19, 22, 23]. The
explanations for such racial disparities in breast cancer survival could take account of stage
at diagnosis, tumor characteristic, access to care/screening, level of treatments and breast
cancer comorbidity [22, 24, 25]. For example, African-American women have been shown
to be more likely than White women to get diagnosed with triple-negative breast cancer-
denoted by the absence of all estrogen receptors (ERs) and progesterone receptor (PR) and
oncogene expression- human epidermal growth factor receptor 2 (HER-2). In the absence
of these three receptors, there are no viable hormonal therapies or therapies targeting HER2
receptors available, thus resulting in reduced survivorship [21, 26-29]. Moreover, African-
American women tend to be diagnosed with more advanced stage of breast cancer
compared to non-Hispanic White women, potentially attributable to the fact that there is
restricted screening or primary care access among African-American women society, or
due to inadequate follow-up clinic visits after an abnormal/inconclusive screening
mammogram result [25, 30, 31]. Notwithstanding that majority race groups (i.e. African
American, Non-Hispanic White, Hispanics) are well studied, relatively little is investigated
within the “other race group”- Asian/Pacific Islander, Non-Hispanic American Indian and
3
Middle Eastern, given limited study population.
Besides the aforementioned factors of demographics and socioeconomic status , hormone
receptor status including ER and PR and oncogene expression- human epidermal growth
factor receptor 2 (HER-2) are also documented as prognostic factors in relation to overall
breast cancer survival [32-34]. When a breast cancer tumor is ‘positive’ for the hormones
estrogen and progesterone, it implies that the tumor is sensitive to hormones thus hormonal
therapy may be very effective on this kind of tumor [32]. Therefore, patients who are
positive for hormone receptors (ER/PR) tend to have higher total survival. Furthermore,
the human epidermal growth factor receptor 2 (HER-2) has also been linked to more
aggressive breast cancers. Up to 20-25% of all breast tumors produce the HER-2 oncogene,
and these patients tend to have higher rates of recurrence and lower overall survival rates
[32, 35, 36]. With the first introduction of HER2-targeted treatment (Trastuzumab), the
survival of HER2 positive breast cancer patients has been gradually improved [37-39]. A
randomized clinical study with over 4,000 female breast cancer participants has revealed
that trastuzumab in addition to standard chemotherapy improved the overall survival (OS)
for HER2-positive breast cancer patients by 37%, and 10-year OS rate also increased from
75.2% to 84% [39]. Accumulative evidence from epidemiologic studies also have
confirmed that breast cancer survival is highly associated with stage at diagnosis: the 5-
year survival for breast cancer tumor confined within the breast is 98.6%, whereas the 5-
year survival for breast cancer found in other parts of the body as well is 25.9% [2, 40]
More genetic factors, on the other hand, are acknowledged as one of the important
4
prognostic determinants of breast cancer survival as well [41-43]. For instance, a
population-based prospective cohort study with1,183 female breast cancer patients
identified that tumor necrosis factor-α (TNF-α) polymorphism (TNF-α
-308
G > A) was
associated with reduced breast-cancer-specific and all-cause mortality [42].Lastly, several
modifiable life-style factors including active smoking are linked with poorer breast cancer
outcome [44-49].
With this large population-based data which has large number of different race/ethnic
groups, we would like to explore the role of demographics, SES, hormone receptors and
tumor characteristics in predicting breast cancer survival, alone and in combination. We
would like to see if the patterns of those determinants hold the same for different tumor
stages. Further, we would like to explore if same predicting patterns hold for race among
each tumor stage groups.
Methods
Study population
Breast cancer cases were collected from the population based cancer registry for Los
Angeles County. The Los Angeles County Cancer Surveillance Program (CSP) is the
population-based cancer registry for Los Angeles County. The CSP is a member of the
statewide population-based cancer surveillance system, the California Cancer Registry
(CCR). It is also part of the National Cancer Institute-funded Surveillance, Epidemiology,
5
and End Results SEER program. The CSP is administered by the University of Southern
California (USC) Keck School of Medicine Norris Comprehensive Cancer Center. [50]
Patients’ demographic (age, race, socioeconomic status) and tumor (stage at diagnosis,
histologic type) characteristics were extracted from the CCR. We excluded patients: 1) who
were male, 2) patients who did not reside in Los Angeles County 3) data not collected
between 2008-2012,since CSP data goes from 2008 to present, we were only using data in
that time frame. After the aforementioned exclusion criteria, there were in total 36,740
subjects in this cohort. However, for the Kaplan–Meier analysis, we further exclude those
who have negative survival time, thus resulting in 36,481 patients. Moreover, for the Cox
proportional hazards regression, besides all the other exclusion criteria, we then excluded
people who were in “other” race/ethnicity and tumor stage was not applicable or error,
yielding 34,705 subject in the study population. According to the CCR variable dictionary,
“error” category here is for those whose tumor stage was unavailable or for whose tumor
stage was apparently no longer localized but no documentation indicating distant
metastasis.
Variable definition
Age was categorized into groups representing study subjects under 40 years old, in 10-year
intervals from 40 to 79 years, and 80 years and older. In order to see more specific survival
patterns in terms of age, we re-categorized age into every 5 years, and the Kaplan-Meier
graph by age groups was plotted accordingly. Original information of patient-level social
economic status (SES) was not assembled by the CCR, so a multi-component measure of
6
SES based on patients' residential census block group at diagnosis was used, incorporating
data on education, occupation, unemployment, household income, poverty, rent, and house
values [51-53]. The initial race variable was derived from several variables including the
patient’s primary race. The final race variable (Non-Hispanic White, Non-Hispanic Black,
Hispanic, Asian/Pacific Islander, Non-Hispanic American Indian, Middle Eastern,
Other/Unknown) that we used for this study was derived from initial race variable and a
Middle Eastern Surname list. Middle Eastern included Afghanistan, Algeria, Armenia,
Egypt, Iraq, Iran, Jordan, Kuwait, Lebanon, Libya, Morocco, Pakistan, Saudi Arabia,
Sudan, Syria, Turkey, Tunisia, and Yemen [54]. Tumor stage was the American Joint
Committee on Cancer (AJCC) stage at diagnosis -I, II, III, IV, unstaged/not applicable or
Error. The stage variable then was re-combined into 3 levels (In situ, stageI&II,
stageIII&IV) for the statistical analysis [55].

Statistical analysis
The distribution of patients/tumor characteristics were first evaluate in general, within each
stage and stage by race, respectively. The Kaplan–Meier method and the log-rank test were
used to show and test differences in survival for all variables and for each sub-group.
Cox proportional hazards regression was used to estimate hazard ratio (HR) and its
corresponding 95% confidence intervals (CI). Univariate analyses were conducted first,
then each independent variables were added in turn to understand their individual effects.
Univariate analysis was performed on the breast cancer survival with each of the six
7
variables (age, race, stage, SES, ER and PR). Then each of the rest of variables was added
into the univariate model that has the smallest AIC value one at a time. Interaction between
the main effects were checked based on the “better model” that was selected previously to
evaluate effect modification. A smaller AIC is considered as a better model fit. But if
difference in AICs between two models is very small (<100), then the more parsimonious
model with similar AIC is selected. We also did the same methods in subsets of this dataset
stratified by stage- stage in situ, stage1&2, and stage3&4 separately.
Survival time was measured in days from the date of diagnosis to the date of death from
breast cancer. Patients who died from other causes were censored at the time of death.
Patients alive at the study end date (12/31/2012) were censored at this time or at the date
of last follow-up. The proportional hazards assumption was checked by statistical testing
of the correlation between weighted Schoenfeld residuals. All the statistical analysis
excluded the missing values. Later, we also re-evaluated all the statistical models with the
missing data to ensure there was no big difference.
All statistical tests were carried out using SAS software version 9.3 (SAS Institute). All P
values reported were two-sided, and those that were <0.05 were considered to be
statistically significant.
Results
Distribution of patient characteristics
There were 36,740 primary diagnoses of female breast cancer occurring in the 5 years from
8
2008 to 2012 recorded in the CSP dataset. Approximately 5% of study population were
under 40 years old, about 18% were between 40-49 years old, and about 50% of the subjects
were between 50-70 years old. 62.7% women got primary diagnosed with stage I or stage
II breast cancer, 14.3% of the patients got diagnosed with stage III or stage IV breast cancer,
only 18% were diagnosed with in situ breast cancer. Non-Hispanic white took up 43.8%
of the participants, 11.8%of the subjects were Non-Hispanic Black, 24.2% were Hispanic,
16% were Asian/Pacific Islander, 0.19% were Non-Hispanic American Indian, 3.83% were
middle Easter.. This study cohort consisted of 14.1% individuals with lowest SES category,
18.0% with lower-middle SES, 18.4% with middle SES class, 23.4% with higher-middle
SES class, and 26.1% with highest SES status. (Table 1)
For Non-Hispanic White patients, the majority of them (83.3%) had early stage breast
cancer (In situ and Stage I&II), and only about 12.8% of the White women had late stage
breast cancer (Stage III&IV). Non-Hispanic Black and Hispanic tended to have more late
stage breast cancer, the proportions were 18.5% and 16.975, relatively. There were about
4.8% patients fell into the “Error” category. (Table 1)
Among all the breast cancer patients, about 76.1% (27,969) were ER positive, 16.1%
(5,903) subjects were ER negative, and approximately 7.68% participants had no/error
information about their ER status. (Table 1)
In addition, there were 63.4%(23,309) people were PR positive, and 25.5% subjects were
PR negative, and only about 10.74% had no/ error information on PR status. (Table 1)

9
Patient distribution based on race by stage
Among 16,091 non-Hispanic White female breast cancer patients, there were 83.3% had
early stage breast cancer (In situ, stage I&II). There were 4,109 non-Hispanic Black female
breast cancer subjects in the dataset, and 75.7% of them had early stage breast cancer, 18.5%
had late stage breast cancer (stage III&IV). Among 8,879 Hispanic patients, 77.6% were
diagnosed with early stage breast cancer, and 17.0% got late stage breast cancer. For 5,878
Asian/Pacific Islander patients, there were about 83% who got early stage breast cancer,
and only 11.8% were diagnosed with late stage breast cancer. The majority of Middle-
eastern women (n=1,406) were diagnosed with early stage breast cancer (81.3%), and less
than 15% of these patients had late stage breast cancer. (Table 2)
For subjects who were older than 50 years old (n= 28,283), there were 13,360 non-Hispanic
White breast cancer patients. Among them, there were 83.4% who got early stage breast
cancer, and only 12.5% got diagnosed with late stage breast cancer. Among 3,241 non-
Hispanic Black, there were 76.8% got diagnosed with early stage breast cancer, and 17.2%
had late stage breast cancer. There were 6,072 Hispanic patients and 79.0% of them had
early stage and 15.2% had late stage breast cancer. Among 4,252 Asian/Pacific Islander,
there were 83.6% had early stage breast cancer and 11.3% had late stage breast cancer. The
majority of the middle-eastern breast cancer patients (n=1,037) had early stage breast
cancer (81.1%), and less than 20% had last stage breast cancer. (Table 2)
In addition, there were 6,637 subjects who were between 40-50 years old. Among the
Whites, there were 84.1% had early stage breast cancer. For Blacks, there were 72.3% had
10
early stage breast cancer and 21.3% late stage breast cancer. There were about 77.0% of
the Hispanics had early stage breast cancer, and 18.4% had the late stage breast cancer. For
Asian/pacific islander and middle-eastern, the distributions were very similar as the general
pattern: majority patients had early stage breast cancer and less than 15% subjects had late
stage breast cancer. (Table 2)
Overall survival by patient characteristics
There were 36,481 in total women in this cohort included in the following statistical
analysis (excluded the missing/error data)-including 16,013 non-Hispanic white, 4,091
non-Hispanic black, 8,830 Hispanic, 5,17 Asian/pacific islanders, 69 non-Hispanic
American Indian, 1,400 Middle Eastern and 261 other/unknown races. The survival
between each race group were statistically significantly different. (Log rank p<0.0001).
The survival patterns of Asian/pacific islander and Middle-eastern were similar and were
better compared to other races, and survival of non-Hispanic black was the worst. The
survival of Hispanic and non-Hispanic white were roughly the same. (Figure 1)
In this cohort, 23,002 patients diagnosed with stage I-II, 5,236 subjects with stage III/IV,
6,576 women had stage in Situ tumor, and 1,667 has unknown stage. The survival
patterns between each stage category were statistically significantly different.(Log rank
p<0.0001). The survival for stage in situ women were the best compared to other stages
and the in situ and Stage I& II survival were fairly similar, not surprisingly, the survival
for stage III&IV were the worst. The “error” group patients had slightly better survival
compared to stage III&IV subjects.(Figure 2) The explanation for the poorer survival
11
pattern of the “Error” group may be that “Error” category actually contained a good
amount of patients with late staged breast cancer but without any documentation, thus
lowering the patients’ survival of this group.
In addition, the survival between each age group was statistically significantly different.
(Log rank p<0.0001). The survival of women who were 45-49 years old were the best
compared to other age categories. Among people between 40-70 years old, as age increased,
breast cancer survival decreased. But all the other age categories among this age range (40-
70 years old) had better survival compared to patients who were less than 40 years old.
Subjects older than 85 years old had the poorest survival (Figure 3)
Moreover, the survival between each ER category was statistically significantly different.
(Log rank p<0.0001). Patients with ER positive status had the best survival compared to
other ER groups. (Figure 4)
The survival between each PR category was statistically significantly different. (Log rank
p<0.0001). Patients with PR positive status had the best survival compared to other PR
status group, and patients with PR ‘test not done’ actually had second to best survival across
all groups. (Figure 4)
Hazard ratios of breast cancer mortality by patient characteristics
Age is statistically significantly associated with breast cancer survival. (p<.0001).
Compared to women age below 40 years old, women aged 40-49, 50-59, 60-69, 70-90 and
older than 80 years were 0.60,0.71, 0.76,1.29, and 3.30 times as likely to die, and p-values
12
were <.0001, <.0001, 0.0006, 0.0018 and <.0001, respectively. (Table 3)
Race is statistically significantly associated with breast cancer survival. (p<.0001).
Compared to non-Hispanic white, the chance of dying for non-Hispanic black, Hispanic,
Asian/pacific islander, non-Hispanic American Indian and middle eastern were 1.61, 1.04,
0.71, 1.08, and 0.81, and p-values for them were <.0001, 0.30, <.0001, 0.83, and 0.02,
respectively. Middle-eastern patients had statistically significant 20% higher chance of
surviving from breast cancer compared to non-Hispanic white. (Table 4)
SES is statistically significantly associated with breast cancer survival. (p<.0001).
compared to women has the highest SES, women had lowest SES, lower-middle SES,
middle SES, and upper-middle SES were 1.92,1.57, 1.47, and 1.20 times as likely to die,
respectively, and all p-values were <0.0005. (Table 5)
Stage is also statistically significantly associated with breast cancer survival. (p<.0001).
Compared to stage in situ, stage I&II and stage III&IV were 2.43 and 12.25 times as likely
to die, respectively. P values were all less than 0.0001. (Table 6)
ER and PR are also statistically significantly associated with breast cancer survival
univariately. (both p<.0001). Compared to ER positive, ER negative and borderline
patients were 2.14 and 1.76 times as likely to die, respectively. P values were <.0001 and
0.11, respectively. (Table 7)
Compared to PR positive, PR negative patients were 2.06 and 1.19 times as likely to die,
respectively. P values were <.0001 and 0.52, respectively. (Table 8)

13
Breast cancer survival adjusted for patient and tumor characteristics
Compared to women with breast cancer who were less than 40 years old, after adjusting
for stage, SES and race, women with breast cancer who were older than 70 years old has
statistically significantly higher chance of dying. Patients between 40-49 years old had the
best survival compared to other age groups including those who were less than 40 years
old. Also, people who are greater than 80 years old were almost 5 times as likely of dying
from breast cancer as those who were less than 40 years old. (p<0.0001) (Table 9).
In addition, after adjusting for age, SES and race, compared to stage in situ, stage I&II and
stage III&IV were 2.23 and 11.06 times as likely to die, respectively. (both p<0.0001).
(Table 9)
Moreover, after adjusting for age, stage and race, compared to the highest SES, women
who were lowest SES, lower-middle SES, middle SES, and upper-middle SES were 1.63,
1.42, 1.38 and 1.20 times as likely to die. (all p values <0.005). (Table 9)
Additionally, after adjusting for age, stage and SES, compared to non-Hispanic white, the
chance of dying for non-Hispanic black, Hispanic, Asian/pacific islander, non-Hispanic
American Indian, middle eastern were 1.28, 0.95, 0.79, 1.13, and 0.85, and p-values for
them were <.0001, 0.25, <.0001, 0.72, and 0.08, respectively. (Table 9).
After adjusting for age, stage, SES and race, compared with patients with ER positive status,
people whose ER results were negative and borderline ER results were 1.54 and 1.38 times
as likely to die from breast cancer, and p values for them were <.0001 and 0.36, respectively.
(Table 9).
14
Lastly, compared to patients with PR positive status, people whose PR status were negative
were 1.44 and 0.90 times as likely to die from breast cancer, after controlling for age, stage,
SES and race. (p<.0001 and 0.7, respectively) The interactions between age and hormone
receptors were statistical significant. (p<0.0001)(Table 9).

Women with primary in situ breast cancer
Compared to women less than 40 years old, women aged 40-49, 50-59, 60-69, 70-90 and
older than 80 years were 0.22, 1.19, 2.15, 5.18, and 13.61 times as likely to die, and p-
values were 0.10, 0.82, 0.30, 0.02 and 0.0003, respectively, after adjusting for SES.
Therefore, only people who were older than 70 years old had statistically significantly
difference in chance of dying from breast cancer compared to women who were less than
40 years old. (Table 10)
Again, after adjusting for age, compared to women has the highest SES, women had lowest
SES, lower-middle SES, middle SES, and upper-middle SES were 2.0, 1.35, 1.67,and 1.65
times as likely to die, and p-values were 0.01, 0.24, 0.03 and 0.03, respectively.(Table 10)

Women with primarily diagnosed stage I&II breast cancer
After adjusting for SES, race, ER and PR, compared to women less than 40 years old,
women aged 40-49, 50-59, 60-69, 70-90 and older than 80 years were 0.77, 0.80, 1.12,
2.25, and 7.39 times as likely to die, and p-values were 0.08, 0.12, 0.43, <0.0001 and
<.0001, respectively. (Table 11)
15
Furthermore, after controlling for age, race, ER and PR, compared to women has the
highest SES, women had lowest SES, lower-middle SES, middle SES, and upper-middle
SES were 1.68, 1.34, 1.26, and 1.11 times as likely to die, and p-values were <.0001, <.0001,
0.001 and 0.10, respectively, after adjusting for age. (Table 11)
Moreover, after adjusting for age, stage and SES, PR and ER, compared to non-Hispanic
white, the chance of dying for non-Hispanic black, Hispanic, Asian/pacific islander, non-
Hispanic American Indian, middle eastern were 1.20, 0.94, 0.77, 1.77, and 0.87, and p-
values for them were 0.01, 0.35, 0.001, 0.16, and 0.31, respectively. (Table 11).
After controlling for age, stage, SES and race, compared with patients with ER positive
status, people whose ER results were negative and borderline ER results were 1.69 and
1.15 times as likely to die from breast cancer, and p values for them were <.0001 and 0.85,
respectively. (Table 11)
Lastly, compared to patients with PR positive status, people whose PR status were negative
and borderline were 1.39 and 0.62 times as likely to die from breast cancer, after controlling
for age, stage, SES and race. (p<.0001 and 29, respectively) (Table 11)

Women with primarily diagnosed stage III&IV breast cancer
After adjusting for race, SES,ER and PR, compared to women less than 40 years old,
women aged 40-49, 50-59, 60-69, 70-90 and older than 80 years were 0.91, 1.14, 1.24,
1.91, and 2.87 times as likely to die, and p-values were 0.39, 0.22, 0.03, <.0001 and <.0001,
respectively. (Table 12)
16
After controlling for age, race, ER and PR, subjects who has lowest SES, lower-middle
SES, middle SES, upper-middle SES were 1.67, 1.54, 1.50, and 1.29 times as likely to die
from breast cancer as patents whose SES level was the highest, and p-values were <.0001,
<.0001, <.0001 and 0.0015, respectively. (Table 12)
Compared to non-Hispanic white, The chance of dying for non-Hispanic black, Hispanic,
Asian/pacific islander, non-Hispanic American Indian, middle-eastern were 1.31, 0.96,
0.86, 0.67 and 0.9, and p-values for them were 0.0001, 0.51, 0.06, 0.48, and 0.42,
respectively, after taking age, SES, ER and PR into account. (Table 12)
After adjusting for age, stage, SES and race, compared with patients with ER positive status,
ER negative patients were 1.45 and 1.51 times as likely to die from breast cancer, and p
values for them were <.0001 and 0.32, respectively. (Table 12)
Finally, compared to patients with PR positive status, people whose PR status were
negative were 1.56 and 1.12 times as likely to die from breast cancer, after controlling for
age, stage, SES and race, and p-values were <.0001 and 0.75, respectively. (Table 12)
To evaluate any missing values would make any difference, every statistical model was
applied to data with all the missing values as well, and we got very similar results as above.

Discussion
In this large, representative series of women diagnosed with breast cancer in California
during recent years, we evaluated the association between demographic and socioeconomic
17
factors, hormone factors and tumor characteristics in relation to female breast cancer
outcome-mortality. We have detected that the survival for women whose tumor was
localized was the best compared to female breast cancer patients with other stages. Survival
was found to increase as SES level increase. This association could not be explained by
other variables in our study since such association was not altered before or after
accounting for other variables. The survival of middle aged women who were between 40-
49 years old were actually the best compared to other age categories Notably, the chance
of dying from breast cancer increased as age increased for women who were greater than
40 years old. Age was a strong predictor of breast cancer survival given that such
association was also independent of other variables in our study. We also demonstrated that
patients with ER/PR positive have better survival chance comparing to ER/PR negative,
independent of other covariates. Interestingly, interactions between age and hormone
receptors were statistically significant. This finding could shed a light in our future studies.
All the above survival patterns stayed similar in each stage subtypes of breast cancer.
Similarity
Findings on SES were in line with the prior reports [13, 15]. Results about age was also
consistent with several previous studies [3-11] -middle aged women have the best breast
cancer survival compared to other age groups and other studies where investigators have
found that younger patients experienced poorer breast cancer survival. Moreover, we found
large racial disparities existed in breast cancer mortality-most apparently, the survival of
non-Hispanic Black women was the poorest compared to other race/ethnicity groups.
18
Several previous studies have demonstrated that breast cancer survival among non-
Hispanic Black was poorer compared to other race/ethnicities [2, 19, 56-61]. Therefore,
our finding with respect to the racial disparity regarding survival among non-Hispanic
Black breast cancer patients was consistent with prior evidence. All the above mentioned
survival patterns stayed similar in each stage subtypes of breast cancer.
Dissimilarity
However, we also revealed that among those with early stage breast cancer, the survival of
Hispanic females was similar to non-Hispanic White females. This result was different
from the finding of several previous reports [16, 62-64] Potential explanations could be
that we lack more information on biologic factors or lifestyle related factors. Breast cancer
tumor growth rate, the likelihood of becoming metastatic, tumor stage at diagnosis and
even the response to available treatment are found to be largely determined by the tumor’s
inherent biologic features [60, 65-67]. For example, Hispanic breast cancer women tend to
have more triple-negative and HER-2 overexpression tumors which may result in more
aggressive breast cancer subtypes [57, 63, 68-70]. Our study lacks information on HER2,
which could be a negative confounder in terms of race and breast cancer mortality. In this
case, HER2 could bias our study in the way of underestimating our hazard ratio comparing
Hispanic women and Non-Hispanic white women. Furthermore, our study did not account
for lifestyle-related factors including smoking status, physical activity, alcohol
consumption and obesity, and these could be important prognostic factors of determining
breast cancer survival [44, 71-77]. For instance, obesity has been identified to be adversely
19
associated with breast cancer survival [76]. Also, higher obesity rate is shown among
Hispanic women compared to non-Hispanic white women[78].Therefore, obesity could
potentially be a confounder in our study.
Strength
Nevertheless, there are a few advantages inherent in our study. With the usage of registry
data, we had less concern about selection bias, especially survival bias comparing to
clinical trials study. For a clinical trial study, patients has to be alive at the beginning of the
study. We also have superiority over clinical studies on generalizability . Problems on
generalizability may arise with the exclusion criteria in a clinical trial study. However, our
study used data from a population-based state-wide registry, thus our study results have
external validity. Additionally, even though, there were numerous studies have been
conducted to evaluate the racial disparities regarding to breast cancer outcomes, very
limited studies have reported the survival of minority groups, especially the Middle-eastern
and Asian/Pacific Islander. Our work extended prior work by using large population-based
data which has very specific race/ethnicity categories to investigate the survival after breast
cancer among different race/ethnicity including Middle-eastern and Asian/Pacific Islander
women who lived in Los Angeles County. Using SEER database with the restriction of
only early staged breast cancer patients, Iqbal et al. reported risk of dying from breast
cancer was higher among African American women than non-Hispanic white women, and
lower in Asian women [61]. Our findings included a more specific race category and all
stage breast cancer patients.
20
We discovered that the survival of Middle-eastern breast cancer patients was similar to
Asian/pacific islander, and the two groups had the better survival compared to other
race/ethnicities. Even after controlling for age at diagnosis, stage, SES, ER and PR, all the
racial disparity still existed. Moreover, after stratifying tumor stage, the pattern of race in
predicting breast cancer survival were very similar in different stage groups, particularly
the “better survival” pattern for Middle Eastern and Asian/Pacific Islanders stayed the same.
The reasons for such difference may be due to the difference in biologic factors and lifestyle
related factors that have been aforementioned, as well as in other sociodemographic aspects
and treatment after breast cancer diagnosis. A women’s awareness and adherence of self or
clinical breast examination and recommended breast mammogram screening could be
influenced by sociodemographic factors, which may also impact the women’s decision in
seeking for proper care after an abnormal result obtained from breast
screening/examination [79-82]. Therefore, by only knowing age, general SES and race may
not be enough, more detailed information are needed including marital status, education
level, insurance status and etc. In addition, subsequent treatment after breast cancer
diagnosis has been suggested to be linked with breast cancer survival, and may be one of
the possible reasons accounting for breast cancer survival disparity [24, 25]. Without
information on subsequent treatment, it is hard to know whether treatment played a role in
attenuating our study results. Therefore, future studies are warranted.
Limitation
There are several relevant limitations needs to be considered when interpreting our study
21
results. Differential misclassification between Hispanics and White women with the
respect of the study outcome-death could be present. Given that death records in a registry
data were obtained from government programs like National death index, such program
are not likely to record an illegal immigrant’s death. There are more illegal immigrants in
Hispanic patients than in white patients. Hence, we might undercount the deaths in
Hispanic patients and not undercount the deaths in White patients. Consequently, such
differential misclassification could bias our study. The result that Hispanic patients had
similar survival pattern compared to non-Hispanic white women could somewhat be
explained by this misclassification. Although, we did analyze our data with existing
information on demographic and tumor characteristics, we still lacked more detailed
information on sociodemographic factors, clinical features, genetic/biomarker aspects and
regimens for treatment arranged to patients. Those factors could be potential confounders
to our study. We also lacked information on screening access. Yet, the benefits and
disadvantages of breast screening program on the survival after breast cancer is still on
debate [24, 83]. Nonetheless, if screening really did help to improve the survival after
breast cancer diagnosis, it might be able to explain the racial disparity in breast cancer
survival, at least partially. For example, African American women tend to be associated
with less likelihood of getting screening mammogram and are more likely to get primarily
late-staged breast cancer diagnosis which will lead to worse survivorship after breast cancer
[84]. Therefore, if we had information on screening in our study that would definitely help
to evaluate whether the race disparity that we found in the study could be explained by
22
screening. Another limitation of study was our inability to control for actual individual-
level measures of SES, instead, we used proxy SES: a multi-component measure of
neighborhood SES based on patients' residential census block group at diagnosis. Although,
such SES should be a good representative of individual-level measures of SES, residual
confounding can still exist which would bias our result in a certain amount. Moreover,
since our race information was based on medical records and patients’ surnames, there
might be differential misclassification in terms of race variable. For example, such
differential misclassification could occur since CCR are more likely to synthesize a
patient’s race from his/her death certificate and those people may be more likely to die.
This could potential help to explain that people with unknown/missing race had high
survival compared to some other certain races. Another example could be that our race
variable was partially based on the Middle-eastern surname. Thus, there might be
differential misclassification in terms of defining Middle-eastern and white. Middle-eastern
people were more likely to be correctly categorized in the Middle-eastern race group
compared to White race category, and Middle-eastern patients tend to have higher survival
rate. Thus, the better survival pattern that we detected for the Middle-eastern female in our
study might be affected by the differential misclassification. Conclusion
In summary, we found an important mortality disparity among non-Hispanic Black,
Hispanics, non-Hispanic Whites, Asian/pacific islander and Middle-eastern. Future
research is warranted to investigate specifically why Middle-eastern and Asian/pacific
islander had better breast cancer survival and whether such difference could be addressed
23
in other ethnicities to improve their breast cancer survival. As a result of our investigation,
we added more strong evidence on social economic status, age and ER/PR as predictors for
breast cancer survival-as the level of SES decreased or age increased or ER-/PR-, the
chance of dying for the female breast cancer patients elevated after adjusting for patients’
other demographic features and tumor characteristics. Overall, being more than 80 years
old, having a diagnosis of stage III or stage IV (late stage disease) breast cancer, having
low SES, being African American and ER/PR negative tumor led to poorest survival
prognosis.

24
References
1. Centers for Disease Control and Prevention, U.S.C.S.W.G., United States Cancer
Statistics: 1999–2014 Incidence and Mortality Web-based Report. . 2017.
2. SEER, U.N.C.I.s.S.E.a.E.R.P., SEER Cancer Statistics Review. 2014.
3. Ibrahim, A., M.A. Salem, and R. Hassan, Outcome of young age at diagnosis of
breast cancer in South Egypt. Gulf J Oncolog, 2014. 1(15): p. 76-83.
4. Cvetanovic, A., et al., Young age and pathological features predict breast cancer
outcome - report from a dual Institution experience in Serbia. J buon, 2015. 20(6):
p. 1407-13.
5. Alieldin, N.H., et al., Age at diagnosis in women with non-metastatic breast cancer:
Is it related to prognosis? J Egypt Natl Canc Inst, 2014. 26(1): p. 23-30.
6. Cluze, C., et al., Analysis of the effect of age on the prognosis of breast cancer.
Breast Cancer Res Treat, 2009. 117(1): p. 121-9.
7. Fernandopulle, S.M., P. Cher-Siangang, and P.H. Tan, Breast carcinoma in women
35 years and younger: a pathological study. Pathology, 2006. 38(3): p. 219-22.
8. Maggard, M.A., et al., Do young breast cancer patients have worse outcomes? J
Surg Res, 2003. 113(1): p. 109-13.
9. Brandt, J., et al., Age at diagnosis in relation to survival following breast cancer: a
cohort study. World J Surg Oncol, 2015. 13: p. 33.
10. Wei, X.Q., et al., Clinical features and survival analysis of very young (age<35)
breast cancer patients. Asian Pac J Cancer Prev, 2013. 14(10): p. 5949-52.
11. V ostakolaei, F.A., et al., Age at diagnosis and breast cancer survival in iran. Int J
Breast Cancer, 2012. 2012: p. 517976.
12. Dreyer, M.S., et al., Socioeconomic status and breast cancer treatment. Breast
Cancer Res Treat, 2017.
13. Sprague, B.L., et al., Socioeconomic status and survival after an invasive breast
cancer diagnosis. Cancer, 2011. 117(7): p. 1542-51.
25
14. Kish, J.K., et al., Racial and ethnic disparities in cancer survival by neighborhood
socioeconomic status in Surveillance, Epidemiology, and End Results (SEER)
Registries. J Natl Cancer Inst Monogr, 2014. 2014(49): p. 236-43.
15. Feinglass, J., N. Rydzewski, and A. Yang, The socioeconomic gradient in all-cause
mortality for women with breast cancer: findings from the 1998 to 2006 National
Cancer Data Base with follow-up through 2011. Ann Epidemiol, 2015. 25(8): p.
549-55.
16. Martinez, M.E., et al., Contribution of clinical and socioeconomic factors to
differences in breast cancer subtype and mortality between Hispanic and non-
Hispanic white women. Breast Cancer Res Treat, 2017.
17. Robin A. Cohen, P.D., and Michael E. Martinez, M.P.H., M.H.S.A. , Health
Insurance Coverage: Early Release of Estimates From the National Health
Interview Survey, January–March 2015. Centers for Disease Control and
Prevention, 2015.
18. Vidal, G., et al., Racial disparities in survival outcomes by breast tumor subtype
among African American women in Memphis, Tennessee. Cancer Med, 2017. 6(7):
p. 1776-1786.
19. Yedjou, C.G., et al., Assessing the Racial and Ethnic Disparities in Breast Cancer
Mortality in the United States. Int J Environ Res Public Health, 2017. 14(5).
20. Eley, J.W., et al., Racial differences in survival from breast cancer. Results of the
National Cancer Institute Black/White Cancer Survival Study. Jama, 1994. 272(12):
p. 947-54.
21. Carey, L.A., et al., Race, breast cancer subtypes, and survival in the Carolina
Breast Cancer Study. Jama, 2006. 295(21): p. 2492-502.
22. Tammemagi, C.M., et al., Comorbidity and survival disparities among black and
white patients with breast cancer. Jama, 2005. 294(14): p. 1765-72.
23. Ries LAG, E.M., Kosary CL, Hankey BF, Miller BA, Clegg L, Mariotto A, Fay MP,
26
Feuer EJ, Edwards BK, editors., SEER Cancer Statistics Review, 1975-2000. 2003.
24. Curtis, E., et al., Racial and ethnic differences in breast cancer survival: how much
is explained by screening, tumor severity, biology, treatment, comorbidities, and
demographics? Cancer, 2008. 112(1): p. 171-80.
25. Silber, J.H., et al., Characteristics associated with differences in survival among
black and white women with breast cancer. Jama, 2013. 310(4): p. 389-97.
26. Morris, G.J., et al., Differences in breast carcinoma characteristics in newly
diagnosed African-American and Caucasian patients: a single-institution
compilation compared with the National Cancer Institute's Surveillance,
Epidemiology, and End Results database. Cancer, 2007. 110(4): p. 876-84.
27. Stark, A., et al., Advanced stages and poorly differentiated grade are associated
with an increased risk of HER2/neu positive breast carcinoma only in White women:
findings from a prospective cohort study of African-American and White-American
women. Breast Cancer Res Treat, 2008. 107(3): p. 405-14.
28. Rakha, E.A., et al., Prognostic markers in triple-negative breast cancer. Cancer,
2007. 109(1): p. 25-32.
29. Bauer, K.R., et al., Descriptive analysis of estrogen receptor (ER)-negative,
progesterone receptor (PR)-negative, and HER2-negative invasive breast cancer,
the so-called triple-negative phenotype: a population-based study from the
California cancer Registry. Cancer, 2007. 109(9): p. 1721-8.
30. Chang, S.W., et al., Racial differences in timeliness of follow-up after abnormal
screening mammography. Cancer, 1996. 78(7): p. 1395-402.
31. Jones, B.A., et al., Inadequate follow-up of abnormal screening mammograms:
findings from the race differences in screening mammography process study
(United States). Cancer Causes Control, 2005. 16(7): p. 809-21.
32. Insitute, U.S.N.C., 2017.
33. Jatoi, I., et al., Breast cancer mortality trends in the United States according to
27
estrogen receptor status and age at diagnosis. J Clin Oncol, 2007. 25(13): p. 1683-
90.
34. Berry, D.A., et al., Estrogen-receptor status and outcomes of modern chemotherapy
for patients with node-positive breast cancer. Jama, 2006. 295(14): p. 1658-67.
35. Slamon, D.J., et al., Human breast cancer: correlation of relapse and survival with
amplification of the HER-2/neu oncogene. Science, 1987. 235(4785): p. 177-82.
36. Slamon, D.J., et al., Studies of the HER-2/neu proto-oncogene in human breast and
ovarian cancer. Science, 1989. 244(4905): p. 707-12.
37. Slamon, D., et al., Adjuvant trastuzumab in HER2-positive breast cancer. N Engl J
Med, 2011. 365(14): p. 1273-83.
38. Goldhirsch, A., et al., 2 years versus 1 year of adjuvant trastuzumab for HER2-
positive breast cancer (HERA): an open-label, randomised controlled trial. Lancet,
2013. 382(9897): p. 1021-8.
39. Perez, E.A., et al., Trastuzumab plus adjuvant chemotherapy for human epidermal
growth factor receptor 2-positive breast cancer: planned joint analysis of overall
survival from NSABP B-31 and NCCTG N9831. J Clin Oncol, 2014. 32(33): p.
3744-52.
40. Clegg, L.X., et al., Cancer survival among US whites and minorities: a SEER
(Surveillance, Epidemiology, and End Results) Program population-based study.
Arch Intern Med, 2002. 162(17): p. 1985-93.
41. Zhang, J., et al., GSTT1, GSTP1, and GSTM1 genetic variants are associated with
survival in previously untreated metastatic breast cancer. Oncotarget, 2017. 8(62):
p. 105905-105914.
42. Duggan, C., et al., Genetic variation in TNFalpha, PP ARgamma, and IRS-1 genes,
and their association with breast-cancer survival in the HEAL cohort. Breast
Cancer Res Treat, 2017.
43. Lu, J., et al., The effect of CYP2D6 *10 polymorphism on adjuvant tamoxifen in
28
Asian breast cancer patients: a meta-analysis. Onco Targets Ther, 2017. 10: p.
5429-5437.
44. Passarelli, M.N., et al., Cigarette Smoking Before and After Breast Cancer
Diagnosis: Mortality From Breast Cancer and Smoking-Related Diseases. J Clin
Oncol, 2016. 34(12): p. 1315-22.
45. Pierce, J.P., et al., Lifetime cigarette smoking and breast cancer prognosis in the
After Breast Cancer Pooling Project. J Natl Cancer Inst, 2014. 106(1): p. djt359.
46. Braithwaite, D., et al., Smoking and survival after breast cancer diagnosis: a
prospective observational study and systematic review. Breast Cancer Res Treat,
2012. 136(2): p. 521-33.
47. Izano, M., et al., Smoking and mortality after breast cancer diagnosis: the health
and functioning in women study. Cancer Med, 2015. 4(2): p. 315-24.
48. Wu, A.H., et al., The California Breast Cancer Survivorship Consortium (CBCSC):
prognostic factors associated with racial/ethnic differences in breast cancer
survival. Cancer Causes Control, 2013. 24(10): p. 1821-36.
49. Nechuta, S., et al., A pooled analysis of post-diagnosis lifestyle factors in
association with late estrogen-receptor-positive breast cancer prognosis. Int J
Cancer, 2016. 138(9): p. 2088-97.
50. (CSP), T.L.A.C.C.S.P., 2017.
51. Tao, L., et al., Breast Cancer Mortality in African-American and Non-Hispanic
White Women by Molecular Subtype and Stage at Diagnosis: A Population-Based
Study. Cancer Epidemiol Biomarkers Prev, 2015. 24(7): p. 1039-45.
52. Yost, K., et al., Socioeconomic status and breast cancer incidence in California for
different race/ethnic groups. Cancer Causes Control, 2001. 12(8): p. 703-11.
53. Yang J, S.C., Harrati A, Clarke C, Keegan THM, Gomez SL., Developing an area-
based socioeconomic measure from American Community Survey data. 2014.
54. Nasseri, K., Construction and validation of a list of common Middle Eastern
29
surnames for epidemiological research. Cancer Detect Prev, 2007. 31(5): p. 424-9.
55. The California Cancer Registry (CCR) Data Dictionary.
56. Li, C.I., K.E. Malone, and J.R. Daling, Differences in breast cancer stage, treatment,
and survival by race and ethnicity. Arch Intern Med, 2003. 163(1): p. 49-56.
57. Ooi, S.L., M.E. Martinez, and C.I. Li, Disparities in breast cancer characteristics
and outcomes by race/ethnicity. Breast Cancer Res Treat, 2011. 127(3): p. 729-38.
58. LeMarchand, L., L.N. Kolonel, and A.M. Nomura, Relationship of ethnicity and
other prognostic factors to breast cancer survival patterns in Hawaii. J Natl Cancer
Inst, 1984. 73(6): p. 1259-65.
59. Newman, L.A., et al., Meta-analysis of survival in African American and white
American patients with breast cancer: ethnicity compared with socioeconomic
status. J Clin Oncol, 2006. 24(9): p. 1342-9.
60. Maskarinec, G., et al., Ethnic differences in breast cancer survival: status and
determinants. Womens Health (Lond), 2011. 7(6): p. 677-87.
61. Iqbal, J., et al., Differences in breast cancer stage at diagnosis and cancer-specific
survival by race and ethnicity in the United States. Jama, 2015. 313(2): p. 165-73.
62. Siegel, R.L., et al., Cancer statistics for Hispanics/Latinos, 2015. CA Cancer J Clin,
2015. 65(6): p. 457-80.
63. Martinez, M.E., et al., Breast cancer among Hispanic and non-Hispanic White
women in Arizona. J Health Care Poor Underserved, 2007. 18(4 Suppl): p. 130-45.
64. Heitz, A.E., et al., Healthy lifestyle impact on breast cancer-specific and all-cause
mortality. Breast Cancer Res Treat, 2017.
65. Cunningham, J.E. and W.M. Butler, Racial disparities in female breast cancer in
South Carolina: clinical evidence for a biological basis. Breast Cancer Res Treat,
2004. 88(2): p. 161-76.
66. Middleton, L.P., et al., Histopathology of breast cancer among African-American
women. Cancer, 2003. 97(1 Suppl): p. 253-7.
30
67. Aziz, H., et al., Early onset of breast carcinoma in African American women with
poor prognostic factors. Am J Clin Oncol, 1999. 22(5): p. 436-40.
68. Chen, L. and C.I. Li, Racial disparities in breast cancer diagnosis and treatment by
hormone receptor and HER2 status. Cancer Epidemiol Biomarkers Prev, 2015.
24(11): p. 1666-72.
69. Howlader, N., et al., US incidence of breast cancer subtypes defined by joint
hormone receptor and HER2 status. J Natl Cancer Inst, 2014. 106(5).
70. Banegas, M.P. and C.I. Li, Breast cancer characteristics and outcomes among
Hispanic Black and Hispanic White women. Breast Cancer Res Treat, 2012. 134(3):
p. 1297-304.
71. Boone, S.D., et al., Active and passive cigarette smoking and mortality among
Hispanic and non-Hispanic white women diagnosed with invasive breast cancer.
Ann Epidemiol, 2015. 25(11): p. 824-31.
72. Holick, C.N., et al., Physical activity and survival after diagnosis of invasive breast
cancer. Cancer Epidemiol Biomarkers Prev, 2008. 17(2): p. 379-86.
73. Irwin, M.L., et al., Influence of pre- and postdiagnosis physical activity on mortality
in breast cancer survivors: the health, eating, activity, and lifestyle study. J Clin
Oncol, 2008. 26(24): p. 3958-64.
74. Kwan, M.L., et al., Alcohol consumption and breast cancer recurrence and survival
among women with early-stage breast cancer: the life after cancer epidemiology
study. J Clin Oncol, 2010. 28(29): p. 4410-6.
75. Kwan, M.L., et al., Postdiagnosis alcohol consumption and breast cancer prognosis
in the after breast cancer pooling project. Cancer Epidemiol Biomarkers Prev, 2013.
22(1): p. 32-41.
76. Dal Maso, L., et al., Effect of obesity and other lifestyle factors on mortality in
women with breast cancer. Int J Cancer, 2008. 123(9): p. 2188-94.
77. Borugian, M.J., et al., Waist-to-hip ratio and breast cancer mortality. Am J
31
Epidemiol, 2003. 158(10): p. 963-8.
78. Hales, C.M., et al., Prevalence of Obesity Among Adults and Youth: United States,
2015-2016. NCHS Data Brief, 2017(288): p. 1-8.
79. Svendsen, R.P., et al., Associations between reporting of cancer alarm symptoms
and socioeconomic and demographic determinants: a population-based, cross-
sectional study. BMC Public Health, 2012. 12: p. 686.
80. Vernon, S.W., et al., Factors associated with perceived risk of breast cancer among
women attending a screening program. Breast Cancer Res Treat, 1993. 28(2): p.
137-44.
81. Aiken, L.S., et al., Perceived determinants of risk for breast cancer and the
relations among objective risk, perceived risk, and screening behavior over time.
Womens Health, 1995. 1(1): p. 27-50.
82. McDonald, P.A., et al., Perceptions and knowledge of breast cancer among African-
American women residing in public housing. Ethn Dis, 1999. 9(1): p. 81-93.
83. Bleyer, A., C. Baines, and A.B. Miller, Impact of screening mammography on
breast cancer mortality. Int J Cancer, 2016. 138(8): p. 2003-12.
84. McCarthy, E.P., et al., Mammography use helps to explain differences in breast
cancer stage at diagnosis between older black and white women. Ann Intern Med,
1998. 128(9): p. 729-36.

32
Tables

Table 1 Demographic and tumor characteristics for women diagnosed with breast
cancer living in Los Angeles County, California, 2008–2012
Characteristics Frequency Percent(%)
Age

age <40 1,820 4.95
40-49 years 6,637 18.1
50-59 years 9,428 25.7
60-69 years 9,033 24.6
70-79 years 6,019 16.4
>=70years 3,803 10.4
Tumor stage

I 13,058 35.5
II 9,993 27.2
III 3,681 10.0
IV 1,573 4.28
In situ 6,612 18.0
Not applicable 47 0.13
Error 1,776 4.83
Race

Non-Hispanic White 16,091 43.8
Non-Hispanic Black 4,109 11.2
Hispanic 8,879 24.2
Asian/Pacific Islander 5,878 16.0
Non-Hispanic American Indian 69 0.19
Middle Eastern 1,406 3.83
Other/Unknown 308 0.84
SES

Lowest SES 4,727 12.9
Lower-middle SES 6,039 16.4
Middle SES 6,183 16.8
Upper-middle SES 7,861 21.4
Highest SES 8,751 23.8
Missing 3,179 8.7
Estrogen Receptor(ER)
Positive/elevated 27,969 76.1
Negative/normal; within normal limits 5,903 16.1
Borderline; undetermined whether positive or negative 50 0.14
Test ordered, results not interpretable 11 0.03
Test ordered, results not in chart 237 0.65
Test not done (test not ordered and not performed) 1,487 4.05
33
Unknown or no information 1,083 2.95
Progesterone Receptor(PR)
Positive/elevated 23,309 63.4
Negative/normal; within normal limits 9,369 25.5
Borderline; undetermined whether positive or negative 116 0.32
Test ordered, results not interpretable 15 0.04
Test ordered, results not in chart 240 0.65
Test not done (test not ordered and not performed) 2,421 6.59
Unknown or no information 1,270 3.46

34
Table 2 Patient distribution based on stage by race in general, within women who
were older than 50 and who were between 40-50 years old for women diagnosed
with breast cancer living in Los Angeles County, California, 2008–2012
Stage Race Frequency Percent(%)
I Non-Hispanic White 6,467 49.5
Non-Hispanic Black 1,223 9.37
Hispanic 2,751 21.1
Asian/Pacific
Islander
1,999 15.3
Non-Hispanic
American Indian
31 0.24
Middle Eastern 509 3.90
Other/Unknown 78 0.60
II Non-Hispanic White 4,175 41.8
Non-Hispanic Black 1,128 11.3
Hispanic 2,646 26.5
Asian/Pacific
Islander
1,594 16.0
Non-Hispanic
American Indian
16 0.16
Middle Eastern 384 3.84
Other/Unknown 50 0.50
III Non-Hispanic White 1,428 38.8
Non-Hispanic Black 504 13.7
Hispanic 1,099 29.9
Asian/Pacific
Islander
493 13.4
Non-Hispanic
American Indian
8 0.22
Middle Eastern 137 3.72
Other/Unknown 12 0.33
IV Non-Hispanic White 634 40.3
Non-Hispanic Black 254 16.2
Hispanic 408 25.9
Asian/Pacific
Islander
203 12.9
Non-Hispanic
American Indian
4 0.25
Middle Eastern 65 4.13
Other/Unknown 5 0.32
In situ Non-Hispanic White 2,753 41.6
Non-Hispanic Black 758 11.5
Hispanic 1,491 22.6
Asian/Pacific
Islander
1,286 19.5
Non-Hispanic 7 0.11
35
American Indian
Middle Eastern 250 3.78
Other/Unknown 67 1.01
Not
applicable/Error
Non-Hispanic White 634 34.8
Non-Hispanic Black 242 13.3
Hispanic 484 26.6
Asian/Pacific
Islander
303 16.6
Non-Hispanic
American Indian
3 0.16
Middle Eastern 61 3.35
Other/Unknown 96 5.27
Stage by race for people older than 50 (N=28,283)
I Non-Hispanic White 5,581 52.1
Non-Hispanic Black 1,032 9.63
Hispanic 2,100 19.6
Asian/Pacific
Islander
1,516 14.2
Non-Hispanic
American Indian
24 0.22
Middle Eastern 407 3.80
Other/Unknown 56 0.52
II Non-Hispanic White 3,366 44.8
Non-Hispanic Black 1,032 13.7
Hispanic 1,661 22.1
Asian/Pacific
Islander
1,116 14.9
Non-Hispanic
American Indian
13 0.17
Middle Eastern 287 3.82
Other/Unknown 35 0.47
III Non-Hispanic White 1,114 43.3
Non-Hispanic Black 365 14.2
Hispanic 656 25.5
Asian/Pacific
Islander
326 12.7
Non-Hispanic
American Indian
3 0.12
Middle Eastern 104 4.04
Other/Unknown 8 0.31
IV Non-Hispanic White 550 45.0
Non-Hispanic Black 191 15.6
Hispanic 266 21.8
Asian/Pacific 154 12.6
36
Islander
Non-Hispanic
American Indian
3 0.25
Middle Eastern 55 4.50
Other/Unknown 4 0.33
In situ Non-Hispanic White 2,195 43.9
Non-Hispanic Black 612 12.2
Hispanic 1,035 20.7
Asian/Pacific
Islander
922 18.4
Non-Hispanic
American Indian
6 0.12
Middle Eastern 184 3.68
Other/Unknown 49 0.98
Not
applicable/Error
Non-Hispanic White 554 38.5
Non-Hispanic Black 195 13.5
Hispanic 354 24.6
Asian/Pacific
Islander
218 15.1
Non-Hispanic
American Indian
2 0.14
Middle Eastern 46 3.19
Other/Unknown 72 5.00
Race by stage for people who were 40-50 years old (N=6637)
I Non-Hispanic White 737 38.4
Non-Hispanic Black 158 8.22
Hispanic 524 27.3
Asian/Pacific
Islander
395 20.6
Non-Hispanic
American Indian
7 0.36
Middle Eastern 83 4.32
Other/Unknown 17 0.88
II Non-Hispanic White 623 31.3
Non-Hispanic Black 214 10.8
Hispanic 695 34.9
Asian/Pacific
Islander
371 18.6
Non-Hispanic
American Indian
3 0.15
Middle Eastern 72 3.62
Other/Unknown 12 0.60
III Non-Hispanic White 232 29.3
Non-Hispanic Black 111 14.0
37
Hispanic 296 37.3
Asian/Pacific
Islander
122 15.4
Non-Hispanic
American Indian
3 0.38
Middle Eastern 27 3.40
Other/Unknown 2 0.25
IV Non-Hispanic White 59 25.3
Non-Hispanic Black 40 17.2
Hispanic 89 38.2
Asian/Pacific
Islander
36 15.5
Non-Hispanic
American Indian
1 0.43
Middle Eastern 8 3.43
Other/Unknown 0 0.00
In situ Non-Hispanic White 486 34.5
Non-Hispanic Black 132 9.38
Hispanic 389 27.6
Asian/Pacific
Islander
329 23.4
Non-Hispanic
American Indian
1 0.07
Middle Eastern 58 4.12
Other/Unknown 13 0.92
Not
applicable/Error
Non-Hispanic White 58 19.8
Non-Hispanic Black 38 13.0
Hispanic 96 32.8
Asian/Pacific
Islander
68 23.2
Non-Hispanic
American Indian
1 0.34
Middle Eastern 14 4.78
Other/Unknown 18 6.14

38
Table 3 Unadjusted Hazard Ratio (95% CI) by age of breast cancer mortality for
women living in Los Angeles County, California, 2008–2012
Age(years) HR 95%CI p-V alues

<40 REFERENCE - -
40-49 0.60 (0.51-0.72) <0.0001
50-59 0.71 (0.61-0.83) <0.0001
60-69 0.76 (0.65-0.89) 0.0016
70-79 1.29 (1.10-1.51) 0.0003
>=80 3.30 (2.83-3.85) <0.0001

Table 4 Unadjusted Hazard Ratio(95% CI) by race of breast cancer mortality for
women living in Los Angeles County, California, 2008–2012
Race HR 95%CI p-V alues
Non-Hispanic White REFERENCE - -
Non-Hispanic Black 1.61 (1.48-1.76) <0.0001
Hispanic 1.04 (0.96-1.13) 0.30
Asian/Pacific Islander 0.71 (0.64-0.79) <0.0001
Non-Hispanic American Indian 1.08 (0.56-2.07) 0.83
Middle Eastern 0.81 (0.67-0.97) 0.02

Table 5 Unadjusted Hazard Ratio(95% CI) by Social Economic Status(SES) of breast
cancer mortality for women living in Los Angeles County, California, 2008–2012
SES HR 95%CI p-V alues

Lowest SES REFERENCE - -

Lower-Middle SES 1.92 (1.74-2.12) <0.0001

Middle SES 1.57 (1.42-1.73) <0.0001

Upper-Middle SES 1.47 (1.33-1.62) <0.0001

Highest SES 1.20 (1.10-1.32) 0.0002

39
Table 6 Unadjusted Hazard Ratio(95% CI) by Tumor Stage of breast cancer mortality
for women living in Los Angeles County, California, 2008–2012
Tumor stage HR 95%CI p-V alues
In situ REFERENCE - -
Stage I&II 2.43 (2.08-2.84) <0.0001
Stage III&IV 12.25 (10.46-14.35) <0.0001

Table 7 Unadjusted Hazard Ratio(95% CI) by Estrogen Receptor status of breast
cancer mortality for women living in Los Angeles County, California, 2008–2012
ER HR 95%CI p-V alues
Positive REFERENCE - -
Negative 2.14 (2.00-2.29) <0.0001
Borderline 1.76 (0.88-3,61) 0.11

Table 8 Unadjusted Hazard Ratio(95% CI) by Progesterone Receptor status of breast
cancer mortality for women living in Los Angeles County, California, 2008–2012
PR HR 95%CI p-V alues
Positive REFERENCE - -
Negative 2.06 (1.94-2.20) <0.0001
Borderline 1.19 (0.70-2.01) 0.52

40
Table 9 Adjusted Hazard Ratio (95% CI) of breast cancer mortality for women living
in Los Angeles County, California, 2008–2012
Characteristics HR 95%CI p-V alues
Age(years) <40 Reference - -

40-49 0.81 (0.68-0.96) 0.015

50-59 0.95 (0.81-1.2) 0.57

60-69 1.14 (0.97-1.34) 0.11

70-79 2.06 (1.75-2.41) <0.0001

>=80 4.98 (4.26-5.83) <0.0001
SES Highest SES Reference - -

Lowest SES 1.63 (1.45-1.82) <0.0001

Lower-Middle SES 1.42 (1.28-1.57) <0.0001

Middle SES 1.38 (1.25-1.57) <0.0001

Upper-Middle SES 1.20 (1.09-1.32) 0.0003
Race Non-Hispanic White Reference - -

Non-Hispanic Black 1.28 (1.16-1.41) <0.0001

Hispanic 0.95 (0.87-1.04) 0.25

Asian/Pacific
Islander
0.79 (0.71-0.88) <0.0001

Non-Hispanic
American Indian
1.13 (0.59-2.17) 0.72

Middle Eastern 0.85 (0.71-1.02) 0.08
Tumor stage In situ Reference - -

Stage I&II 2.23 (1.90-2.61) <0.0001

Stage III&IV 11.06 (9.44-12.97) <0.0001
ER Positive Reference - -
Negative 1.54 (1.40-1.69) <0.0001
Borderline 1.38 (0.69-2.78) 0.36
PR Positive Reference - -
Negative 1.44 (1.32-1.57) <0.0001
Borderline 0.90 (0.53-1.53) 0.70
*p values <0.0001 for both interactions between age and ER, age and PR

41
Table 10 Adjusted Hazard Ratio(95% CI) of breast cancer mortality for women with
primary In situ breast cancer living in Los Angeles County, California, 2008–2012
Characteristics HR 95%CI p-V alues
Age(years) <40 Reference - -

40-49 0.22 (0.04-1.34) 0.10

50-59 1.19 (0.34-2.73) 0.82

60-69 2.15 (0.62-4.72) 0.30

70-79 5.18 (1.38-10.32) 0.02

>=80 13.61 (4.30-31.94) 0.0003
SES Highest SES Reference - -

Lowest SES 2.0 (1.18-3.36) 0.01

Lower-Middle SES 1.35 (0.82-2.22) 0.24

Middle SES 1.67 (1.04-2.69) 0.03

Upper-Middle SES 1.65 (1.04-2.60) 0.03

42
Table 11 Adjusted Hazard Ratio(95% CI) of breast cancer mortality for women with
primarily diagnosed stage I&II breast cancer living in Los Angeles County,
California, 2008–2012
Characteristics HR 95%CI p-
Values
Age(years) <40 Reference - -

40-49 0.77 (0.58-1.04) 0.08

50-59 0.80 (0.61-1.06) 0.12

60-69 1.12 (0.85-1.47) 0.43

70-79 2.25 (1.72-2.94) <0.0001

>=80 7.39 (5.67-9.61) <0.0001
SES Highest SES Reference - -

Lowest SES 1.68 (1.43-1.96) <0.0001

Lower-Middle SES 1.34 (1.16-1.54) <0.0001

Middle SES 1.26 (1.10-1.45) 0.001

Upper-Middle SES 1.11 (0.98-1.27) 0.10
Race Non-Hispanic
White
Reference - -
Non-Hispanic
Black
1.20 (1.04-1.38) 0.01
Hispanic 0.94 (0.83-1.07) 0.35
Asian/Pacific
Islander
0.77 (0.67-0.90) 0.001
Non-Hispanic
American Indian
1.77 (0.79-3.95) 0.16
Middle Eastern 0.87 (0.67-1.14) 0.31
ER Positive Reference - -
Negative 1.69 (1.46-1.95) <0.0001
Borderline 1.15 (0.29-4.62) 0.85
PR Positive Reference - -
Negative 1.39 (1.22-1.58) <0.0001
Borderline 0.62 (0.26-1.50) 0.29

43
Table 12 Adjusted Hazard Ratio(95% CI) of breast cancer mortality for women with
primarily diagnosed stage III&IV breast cancer living in Los Angeles County,
California, 2008–2012
Characteristics HR 95%CI p-V alues
Age(years) <40 Reference - -

40-49 0.91 (0.73-1.13) 0.39

50-59 1.14 (0.93-1.39) 0.22

60-69 1.24 (1.01-1.52) 0.03

70-79 1.91 (1.55-2.36) <0.0001

>=80 2.87 (2.32-3.55) <0.0001
SES Highest SES Reference - -
Lowest SES 1.67 (1.41-1.98) <0.0001
Lower-Middle SES 1.54 (1.31-1.80) <0.0001
Middle SES 1.50 (1.29-1.76) <0.0001
Upper-Middle SES 1.29 (1.10-1.51) 0.0015
Race Non-Hispanic
White
Reference - -

Non-Hispanic Black 1.31 (1.14-1.51) 0.0001

Hispanic 0.96 (0.84-1.09) 0.51

Asian/Pacific
Islander
0.86 (0.73-1.01) 0.06

Non-Hispanic
American Indian
0.67 (0.21-2.08) 0.48

Middle Eastern 0.90 (0.69-1.17) 0.42
ER Positive Reference - -
Negative 1.45 (1.27-1.65) <0.0001
Borderline 1.51 (0.67-3.40) 0.32
PR Positive Reference - -
Negative 1.56 (1.37-1.76) <0.0001
Borderline 1.12 (0.56-2.26) 0.75

44
Figures

Figure 1: K-P survival plot: Overall survival of women diagnosed with breast cancer living
in Los Angeles County by race, 2008–2012
*1: non-Hispanic White; 2: non-Hispanic Black; 3: Hispanic; 4: Asian/Pacific Islander; 5: Non-Hispanic American Indian; 6: Middle-eastern; 9:
Other/unknow

45

Figure 2: K-P survival plot: Overall survival of women diagnosed with breast cancer living
in Los Angeles County by stage, 2008–2012

46

Figure 3: K-P survival plot: Overall survival of women diagnosed with breast cancer living
in Los Angeles County by age categories, 2008–2012

47

Figure 4: K-P survival plot: Overall survival of women diagnosed with breast cancer living
in Los Angeles County by ER status, 2008–2012

48

Figure 5: K-P survival plot: Overall survival of women diagnosed with breast cancer living
in Los Angeles County by PR status, 2008–2012

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

Disparities in colorectal cancer survival among Latinos in California

PDF

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

PDF

Risk factors and survival outcome in childhood alveolar soft part sarcoma among patients in the Children’s Oncology Group (COG) Phase 3 study ARST0332

PDF

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

PDF

Air pollution and breast cancer survival in California teachers: using address histories and individual-level data

PDF

Disparities in gallbladder, intra-hepatic bile duct, and other biliary cancers among multi-ethnic populations: a California Cancer Registry study

PDF

Analysis of factors associated with breast cancer using machine learning techniques

PDF

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

PDF

The role of pesticide exposure in breast cancer

PDF

Construction of a surgical survival prediction model of stage IV NSCLC patients-based on seer database

PDF

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

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

A novel risk-based treatment strategy evaluated in pediatric head and neck non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) patients: a survival analysis from the Children's Oncology Group study...

PDF

The impact of the COVID-19 pandemic on cancer care delivery

PDF

Incidence and survival rates of the three major histologies of renal cell carcinoma

PDF

Body size and the risk of prostate cancer in the multiethnic cohort

PDF

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

PDF

Visual acuity outcomes after cataract extraction in Chinese Americans: the Chinese American Eye Study (CHES)

Asset Metadata

Creator Yu, Jing (author)

Core Title Predictive factors of breast cancer survival: a population-based study

School Keck School of Medicine

Degree Master of Science

Degree Program Applied Biostatistics and Epidemiology

Publication Date 02/07/2018

Defense Date 02/07/2018

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

Tag breast cancer,mortality,OAI-PMH Harvest,predictive factors,Survival

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cockburn, Myles (committee chair), Franklin, Meredith (committee member), Mckean-Cowdin, Roberta (committee member)

Creator Email ave688@gmail.com,jyu703@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-469025

Unique identifier UC11268425

Identifier etd-YuJing-6007.pdf (filename),usctheses-c40-469025 (legacy record id)

Legacy Identifier etd-YuJing-6007.pdf

Dmrecord 469025

Document Type Thesis

Rights Yu, Jing

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