University of Southern California Dissertations and Theses

Prostate cancer disparities among Californian Latinos by country of origin: clinical characteristics, incidence, treatment received and survival

(USC Thesis Other)

Prostate cancer disparities among Californian Latinos by country of origin: clinical characteristics, incidence, treatment received and survival

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

PROSTATE CANCER DISPARITIES AMONG CALIFORNIAN LATINOS BY COUNTRY
OF ORIGIN: CLINICAL CHARACTERISTICS, INCIDENCE, TREATMENT RECEIVED
AND SURVIVAL

By

Alexis Ruth Gaines, MSc

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
MOLECULAR EPIDEMIOLOGY

December 2017

Copyright: 2017 Alexis Ruth Gaines, MSc

Dedication

I dedicate this document to my loving husband, Dr. Stephen J. Freedland, my amazing parents,
Mrs. Edith M. Young and Mr. Mosely C. Gaines, Jr., and my host of family and friends. They
have loved me and stood by me, and without their support, I wouldn’t be where I am today.
Thanks for everything.

I would also like to dedicate this to the men and women who work tirelessly for the California
Cancer Registry to collect, clean, organize, and maintain these data from which my project is
developed. For that, I thank you.

Acknowledgments

This work would not have come to fruition without the guidance and assistance of so many
people. To each of you, I am sincerely grateful for the time you have devoted to teaching,
guiding and supporting me, but most importantly, helping me to achieve my dreams.

I was incredibly fortunate and honored to have my thesis committee led by Dr. Mariana C. Stern,
who not only inspired me to chase my dreams, but pushed me to generate the best work I could
for this document and my future career. Without her unwavering support, guidance, patience,
and encouragement through my pursuit of a second Masters degree, I would have taken a more
circuitous route to my doctorate degree. Her words and always being in my corner have helped
me as a researcher and a student, and I hope I am able to inspire future students the way she has
inspired me. Thank you so much for everything.

To my wonderful thesis committee, Dr. Ann Hamilton and Dr. Lihua Liu, I could not have asked
for a more inspiring group of women to work with. Thank you for the support.

Additionally, I have been fortunate enough to have several collaborators who have dedicated
their time and efforts to review this work, provide insight, guidance and be a sounding board for
a developing researcher like me. Without you and your varied expertise, we would not be able to
add this contribution to the literature and gain more insight into prostate cancer in the American-

based Latino community. Thus, in no specific order, I’d like to say thank you to the following:
Dr. Andre E. Kim, Dr. Dennis Deapen, and Ms. Juanjuan Zhang.

Also, in no specific order, I would like to thank Dr. Victoria Cortessis (Preventive Medicine), Dr.
Farzana Choudhury (Ophthalmology), and Dr. Joseph Landolph (Molecular Microbiology and
Immunology). Class with all of you both inspired and challenged me, and for that I'm forever
grateful. I appreciate all of you for believing in me, so having you in my corner was and is an
honor. The time I spent with each of you helped to shape my graduate career and help me
cultivate my passions.

All of the aforementioned are amazing educators and I thank you for all you do.

Table of Contents

DEDICATION

ACKNOWLEDGMENTS

LIST OF TABLES

LIST OF ABBREVIATIONS

ABSTRACT ............................................................................................................ …………1

1. INTRODUCTION ................................................................................................................. 5

1.1. PROSTATE CANCER AND DISEASE BURDEN. ............................................................. 5

1.2. RISK FACTORS ........................................................................................................ 6

1.2.1. HEALTH DISPARITIES ................................................................................... 7

1.2.2. RACE AND ETHNICITY .................................................................................. 8

1.2.3. PROSTATE CANCER SCREENING AND ACCESS TO CARE ............................... 10

1.3. PROSTATE CANCER AND LATINOS ......................................................................... 12

1.4. CALIFORNIA CANCER REGISTRY ............................................................................ 15

1.5. RESEARCH OBJECTIVES ......................................................................................... 15

2. METHODS AND STUDY DESIGN ....................................................................................... 17

2.1. CASE IDENTIFICATION .......................................................................................... 17

2.2. STUDY VARIABLES ............................................................................................... 18

2.3. DATA ANALYSIS .................................................................................................. 22

3. RESULTS ......................................................................................................................... 22

3.1. PATIENT CHARACTERISTICS ................................................................................. 22

3.2. CLINICAL CHARACTERISTICS AND TREATMENT RECEIVED ................................... 24

3.3. IMPACT OF NATIVITY ON PATIENT AND CLINICAL CHARACTERISTICS AND TREATMENT
RECEIVED .................................................................................................................. 27

3.4. PROSTATE CANCER SURVIVAL AMONG CALIFORNIAN LATINOS ............................ 29

4. DISCUSSION .................................................................................................................... 32

TABLES ............................................................................................................................... 39

REFERENCES ....................................................................................................................... 54

List of Tables

TABLE 1A – PATIENT CHARACTERISTICS OF PROSTATE CANCER CASES DIAGNOSED BETWEEN
1995 AND 2012 IN CALIFORNIA BY RACE/ETHNICITY (N = 321,433) ................... …………39

TABLE 1B – PATIENT CHARACTERISTICS OF LATINO PROSTATE CANCER CASES DIAGNOSED
BETWEEN 1995 AND 2012 IN CALIFORNIA BY COUNTRY OF ORIGIN (N = 51,266) ....... 40

TABLE 2A – CLINICAL CHARACTERISTICS OF PROSTATE CANCER CASES DIAGNOSED BETWEEN
1995 AND 2012 IN CALIFORNIA BY RACE/ETHNICITY (N = 321,433) ................................... 41

TABLE 2B – CLINICAL CHARACTERISTICS OF LATINO PROSTATE CANCER CASES DIAGNOSED
BETWEEN 1995 AND 2012 IN CALIFORNIA BY COUNTRY OF ORIGIN (N = 51,266) ....... 44

TABLE 3A – PATIENT CHARACTERISTICS OF LATINOS BY NATIVITY STATUS (N = 51,266)… 47

TABLE 3B – CLINICAL CHARACTERISTICS OF LATINOS BY NATIVITY STATUS (N = 51,266) .. 48

TABLE 4A – CRUDE AND MULTIVARIABLE COX PROPORTIONAL HAZARDS MODELS OF
ETHNICITY AND COUNTRY OF ORIGIN ................................................................................. 51

TABLE 4B – CRUDE COX PROPORTIONAL HAZARDS MODELS OF ETHNICITY AND COUNTRY OF
ORIGIN STRATIFIED BY NATIVITY STATUS .......................................................................... 52

TABLE 4C –MULTIVARIABLE COX PROPORTIONAL HAZARDS MODELS OF ETHNICITY AND
COUNTRY OF ORIGIN STRATIFIED BY NATIVITY STATUS ...................................................... 53

List of Abbreviations
AUA: American Urological Association
CI: confidence interval
CCR: California Cancer Registry
COO: country of origin
CSA: Central/South American
IQR: interquartile range
HR: hazard ratio
NAACCR: North American Association of Central Cancer Registry
NHIA: NAACCR Hispanic Identification Algorithm
NLB: Non-Latino Black
NLW: Non-Latino White
NOS: not otherwise specified
P: p-value
PC: Prostate cancer
PSA: Prostate specific antigen (measured in ng/mL)
PPV: positive predictive value
SES: socioeconomic status
SEER: Surveillance, Epidemiology and End Results
US: United States
USPSTF: United States Preventive Services Task Force

1
Abstract

In California, much like the rest of the United States (US), Prostate Cancer (PC) is the most
common non-skin cancer among men. However, PC incidence patterns, clinical characteristics,
treatment received, and survival are poorly characterized among US Latinos. Latinos comprise a
significantly heterogeneous population with unique national identities, demographic
characteristics, and ancestry. Since access to care, treatment received and genetic factors may
affect prostate risk and survival, the racial admixture among differing ethnic populations may
impact PC risk and survival among Latino subpopulations. However little is known about how
prostate risk and survival differ among the Latino subpopulations based on their country of
origin (COO). As such, we investigated differences in demographic and clinical characteristics
related to incidence and survival of 321,433 men diagnosed with PC in California from 1995-
2012 accounting for ethnicity and COO using data from the California Cancer Registry (CCR).

Determination of Latino ancestry and COO were identified using the North American
Association of Central Cancer Registry (NAACCR) Hispanic Identification Algorithm (NHIA).
We grouped Latinos (15.9% of all cases) into the following COO: Mexico (26.3%),
Central/South America (CSA, 6.9%), Cuba (1.3%), Puerto Rico (0.8%), and not otherwise
specified (NOS, 64.6%). Non-Latino white (NLW, 73.4%) and non-Latino Black (NLB, 10.4%)
Americans were used for comparison. Nativity (US-born vs. foreign-born) was defined based on
birthplace information available in the CCR records, of which less than 5% among all ethnicities

2
were unknown. Socioeconomic status (SES) was based on the geocoding of the participant's
residential address at diagnosis at the census block group level, using a well-established
methodology. Frequencies were calculated for patient and clinical characteristics of men with
incident prostate cancer by race/ethnicity, Latino subpopulation and nativity status. Patient
characteristics included age, SES, insurance status, vital status and marital status. Clinical
characteristics included year of diagnosis, Prostate Specific Antigen test (PSA) status and results,
number of cores taken on prostate biopsy, of the biopsy cores taken, the number that were
positive for PC, pathological stage at diagnosis, Gleason score at diagnosis and after surgery,
time from diagnosis to treatment, treatment status, treatment type, tumor size, lymphadenectomy
status, and survival time after diagnosis. Pearson’s Chi-square was used to test differences
among ethnicities and subgroups. Univariable and multivariable Cox proportional hazards ratios
were used to estimate survival in five-year increments by nativity status, ethnicity and Latino
subpopulation.

Incidence related: Latinos were older at diagnosis than NLB, with CSA having the greatest
proportion of diagnoses under age 50. Latinos were more likely to reside in lower SES areas than
NLW and more likely to be uninsured. NLB were more likely to have elevated PSA (94.2%)
than NLW (89.0%) and Latinos (91.8%, p<0.001). Latinos and NLB had the highest proportion
of non-localized PC relative to NLW (18.3% and 18.1% vs. 15.6%, p<0.001), among them
Mexicans having the highest proportion (22.5%). Among all Latinos by nativity status, foreign-

3
born were older, mainly came from Mexico, of low SES and were uninsured. Foreign-born
Latinos had a higher percentage of non-localized PC, and more cores positive for PC on biopsy.

Treatment related: Mexicans also had the longest time from diagnosis to treatment relative to
other Latino subgroups. Among all men, surgery and radiation were the most used treatments
followed by no treatment. Cubans had the lowest receipt of active treatment. Fewer Latinos
reported receiving radiation therapy (28.5%) than NLB or NLW (31.8% and 32.2%, p<0.001),
among them Puerto Ricans had the highest proportion of radiation treatment whereas CSA had a
greater proportion of surgery. Latinos had greater proportion of cases with no treatment reported
(15.3%) relative to NLW and NLB. Foreign-born Latinos had more cases treated with radiation.

Survival related: On crude analysis, NLB had greater risk of death (HR 1.15, 95% CI 1.12-1.17,
p<0.001) than NLW. There were no differences in survival between Latinos (as a group) and
NLW (p=0.32). On multivariable analysis, NLB had similar results (HR 1.19, 95% CI 1.17-1.22,
p<0.001), however Latino ethnicity was associated with lower risk of death than NLW after
adjusting for various patient and clinical characteristics (HR 0.87, 95% CI 0.85-0.89, p<0.001).
Among only Latino subpopulations, relative to NLW, on crude analysis, Mexicans and Cubans
had a 12% and 18% increased risk of death (Mexican: HR 1.12, 95% CI 1.08 – 1.15, Cuban: HR
1.18, 95% CI 1.05 – 1.33, all p<0.001) while both CSA and NOS all had 21% and 4%,
respectively, decreased risk of death (CSA: 0.79, 95% CI 0.74 – 0.85, NOS: HR 0.96, 95% CI

4
0.94 – 0.98, all p<0.001). On both crude and multivariable analysis, there were no differences
between NLW and Puerto Ricans (p=0.67), however only on multivariable analysis was there no
difference between NLW and Cubans (p=0.45). Multivariable adjustment resulted in a 4%
decreased risk of death for Mexicans (HR 0.96, 95% CI 0.93 – 1.00, p=0.03), however, for the
results little changed after multivariable adjustment for the other COOs. Similar patterns of
results were observed when crude and multivariable models were analyzed by nativity status.

We observed differences in PC incidence patterns, treatment received, and survival among
Latino men in California according to their ethnicity and COO. This highlights the importance of
considering the heterogeneity in this minority population in understanding the cancer
determinants, patterns of care, and survival for PC among Latinos.

5
1. Introduction
1.1. Prostate Cancer and Disease Burden
Prostate cancer (PC) is a slow growing, malignant adenocarcinoma, which is typically
asymptomatic until it reaches advanced stages. A disease of the aged, PC is the most commonly
diagnosed cancer among men in the developed world, after skin cancer, and is the leading cause
of cancer-related death as well.
1
According to the Global Burden of Disease Cancer
Collaboration, in 2015, there was an estimated 1.6 million incident cases of PC and 366,000 PC
deaths worldwide.
2
In the United States (US) specifically, it is estimated that as of 2017, there
will be 161,360 newly diagnosed cases of PC and 26,730 PC deaths.
3
Globally, as of 2015, age-
standardized PC incidence and death rates were lowest in South and Eastern Asia and highest in
Australasia, North America and Western Europe.
2
Among people of Hispanic descent across
Latin America, there is a wide range of PC incidence and mortality rates by country of origin
(COO).
4

Approximately 66% of the increase in incident PC cases in the US from the early 1990s to the
2010s can be attributed to an increase in life expectancies, the introduction of PSA testing for
screening, and a steadily growing population.
5, 6
Furthermore, though not completely consistent,
some data suggest that the PC burden is disproportionately distributed among developed
countries relative to developing, which in part may be attributable to greater access to Prostate

6
Specific Antigen (PSA) screening.
7-10
It is known that as people move from countries with “low”
PC incidence rates to countries with “high” incidence rates their risk of PC goes up.
11
It is also
known that within regions of “low” incidence, different countries have different PC risk.
12

1.2. Risk Factors
PC is a multifactorial and complex disease, the etiology of which is not well understood. The
strongest risk factors identified to date are increased age, PC family history, African ancestry,
and key susceptibility alleles.
13-17
The role of other putative risk factors such as high body mass
index
18
and selected dietary factors is less conclusive.
19, 20
The relevance of environmental risk
factors on PC risk is highlighted by studies that have shown that individuals who migrate from
low-incidence to higher-incidence regions develop similar risk patterns to natives in as quickly
as one generation.
21, 22
Often, there are environmental differences between individuals who were
born and raised in one country and their descendants who have migrated, as well as among
individuals of different races and ethnicities within one nation’s bounds. Differing exposures,
lifestyles, traditions, genetic heritage, and behaviors can all contribute to disease incidence
patterns. In addition, differences in screening behaviors may also contribute to differences in
incidence patterns.

7
1.2.1. Health Disparities
Health status disparities refer to differences in disease incidence, mortality, outcomes, and
disease prevalence among subgroups defined by factors such as geography, culture, gender,
sexuality, race/ethnicity, or socioeconomic status (SES).
23-25
Not inclusively, health disparities
may be due to differences in biology, health status, quality, and access to care, often the result of
measurable inequalities in health care services received or rendered, as well as exposure to
environmental risk factors that contribute to disease development. Often, it is difficult to
disentangle how much of the disparity among ethnic groups is due to biology and genetics,
perceived discrimination, poor health seeking behaviors, access to care, quality of care received,
SES, or other factors. For example, historically, Non-Latino Whites (NLW) and Non-Latino
Blacks (NLB) have had higher PC incidence (20% and 200%, respectively) and mortality (25%
and 300%, respectively) than their US Latino counterparts.
26
Moreover, relative to NLW, NLB
have worse outcomes after treatment, including increased risk of biochemical (i.e. PSA)
recurrence and metastasis
27
, higher likelihood of an abnormal PSA screen on initial biopsy
28
, and
have more aggressive disease at diagnosis.
29
Furthermore, some studies have shown that, even
when controlling for access to care, differences in SES, and clinical features, such as body mass
index, age, and PSA, NLB are still twice as likely to be diagnosed with PC on initial diagnosis
than NLW, suggesting that these disparities may be due to genetic susceptibility.
30

8
US Latinos, on the other hand, have been shown to have lower incidence, but still show marked
disparities in terms of outcomes and treatment patterns relative to NLW. Relevant to risk
stratification and biology, US Latinos were also shown to need lower optimal cutoff values for
PSA testing relative to NLB and NLW.
31, 32
Interestingly, US Latinos also have approximately
20% of their population below the poverty line, fewer high school graduates, higher percentages
of individuals under 65 years old with no health care coverage, no access to regular medical care
26, 33
and presented with more advanced PC on diagnosis than NLB and NLW.
34
As such, it is
critical to understand how health disparities, such as differences in disease aggression and access
to care, present in various populations and whether they impact treatment selections, prognosis,
and survival.

1.2.2. Race and Ethnicity
Though its exact definition is unclear, the term “race” is frequently used to describe a social and
hierarchical construct that separates people of similar origins based upon variations in skin color.
These skin color-based groups help individuals group themselves and others based on cultural,
religious, and political identities.
35, 36
Race differs from ethnicity, which refers to groups of
people from a common ancestral nation, region, or culture. While race has historically been
evaluated by self-report, contemporary technologies have allowed researchers to better
understand genetic ancestry after genetic testing has been performed to confirm ethnic heritage.

9
Therefore, within a given ethnic group there could be individuals of various races, and
commonly, individuals with admixed ancestral origins.

Understanding the relationship between race and cancer is complex at best, due to many factors
associated with racial groups including socioeconomic status, access to care, and genetic
ancestry. Many US-based studies have examined differences in PC patterns between NLB and
NLW men because NLB have the highest PC incidence (NLB: 198.4 per 100,000 vs. NLW:
114.8 per 100,000 vs. Latinos: 104.9 per 100,000) and mortality rates (NLB: 42.8 per 100,000
vs. NLW: 18.7 per 100,000 vs. Latinos: 16.5 per 100,000) among all racial/ethnic groups.
30,38

Latino men have been largely unstudied and this might be in part because their PC risk is similar,
but slightly lower than NLW men. However, Latinos are heterogeneous in ancestry and COO,
and differences in PC incidence and mortality within the Latino subgroups may provide insights
into the etiology of this disease. Furthermore, identifying Latino subgroups at higher risk may
assist in targeting groups who might benefit from closer screening.
39

Thus, in order to understand PC disparities associated with racial groups, more studies are
required to evaluate the role of genetic ancestry, social factors, and determinants of differences in
incidence, treatment received, and outcomes in attempts to reduce incidence and improve
outcomes.

10
1.2.3. Prostate Cancer Screening and Access to Care
Access to care may affect stage at diagnosis, screening history, and treatment selection. These
factors, in turn, may affect survival. Access to care encapsulates more than whether or not a
patient receives the right care at the right time for the right health condition
40-42
; it also includes
access to screening for non-communicable, communicable, and chronic illnesses, receiving basic
preventive health screenings, having the ability to utilize healthcare system and reap benefits,
physically (i.e. distance to care facilities), financially (i.e. insurance status, qualifications for
care), and knowingly (i.e. having the ability to make informed decisions about their health).
Access to care also refers to whether or not gatekeepers are available to help patients maneuver
within the system to get the care, treatments, and tests they need. Depending on the health
system model, these gatekeepers could include primary care physicians (in health maintenance
organization and universal health care models) or health system intermediaries. Closing gaps in
access to care disparities includes determining where the system-wide deficiencies lie. Research
has shown that patients with lower SES are less likely to come in contact with the health care
system in general.
43
They have less access to primary care providers and specialists, less
education to know how to make informed decisions
44
and maneuver within the system once they
encounter it, and overall worse outcomes.
45

Relevant to PC, in 1986, the introduction of the biomarker based PSA blood test for screening
resulted in a widespread, dramatic increase in incidence rates for PC among all racial ethnic

11
groups.
46
The PSA test measures the amount of protein generated by the prostate in whole blood.
Typically elevated in men with PC, benign prostatic hyperplasia, and prostatitis, the PSA test is
most commonly used in conjunction with digital rectal examinations to detect PC. Measured in
units of nanograms per milliliter (ng/mL), the PSA test lacks the ability to differentiate between
PSA increases generated from aggressive cancers from those indicative of indolent cancers that
will never progress, or the aforementioned benign conditions. The goal of screening is to
identify, among all men, who is at greatest risk of PC death and benefit the most from further
diagnostic procedures. In recent years, the use of PSA testing has been met with controversy
regarding whether the risk of over-diagnosis and overtreatment
47
outweigh the benefits of early
detection. The US Preventive Task Force (USPSTF) and American Urological Associations
(AUA) both recommend against screening for men at “average risk”. Specifically, the USPSTF
has stated that the risks of PSA screening outweigh the harms
48
, whereas the AUA recommends
PSA testing in the context of shared decision-making for men between the ages of 55 and 69.
49

Definitions of “average risk” are mixed, and it’s unclear how different racial/ethnic groups fit
into this. Whereas the introduction of PSA testing has clearly impacted incidence among all men,
results from the Prostate, Lung, Colorectal and Ovarian Cancer Screening Trial suggests that the
introduction of PSA tests has had little to no impact on mortality among US men.
50

For Latinos, specifically, the impact of PSA testing on incidence has historically followed the
same trends as NLB and NLW. It has been suggested that the differences in incidence rates

12
between NLW and Latinos is due to lower receipt of PSA screening among Latino men.
51

Relative to NLW, Latino men are also less likely to be diagnosed with Stage I (localized) PC,
which provides some circumstantial evidence that Latino men may have lower receipt of PSA
testing. In the PSA era, most PCs are diagnosed in its early stages.
52
However, two studies have
shown that receipt of a PSA test is tied to SES and that more affluent men are more likely to
undergo PC screening.
53, 54
According to 2014 US Census estimates from the Current Population
Survey Annual Social and Economic Supplement questionnaire, 23.6% of individuals below the
poverty line were Latino, relative to 10.1% of NLW and 26.2% of NLB.
55
The relationship
between SES, a proxy for income and poverty, and cancer is complex. Interestingly, some data
have suggested that while Latinos of lower SES have better outcomes after cancer diagnoses
compared to their NLB and NLW counterparts. This phenomenon, coined the “Latino paradox”,
is still not fully understood.
56, 57
Specifically for PC, studies of SES and race/ethnicity have
shown that immigrant Latinos have better survival after PC compared to US-born Latinos,
especially among those living within ethnically enriched neighborhoods (“enclaves”).
58
As such,
understanding the relationship among race, ethnicity and SES will help us have better insight
into how these factors may impact access to care, cancer risk, patterns of care, and mortality.
59

1.3. Prostate Cancer and Latinos
According to the US Census Office of Management and Budget, a person is defined as
“Hispanic” or “Latino” if their origins are from Mexico, Cuba, Puerto Rico, Central/South

13
America (CSA), or Spain, irrespective of the race (or races) a person identifies with.
60
2016
census estimates showed that there were 56.5 million Latinos in the US comprising 17.6% of the
total population, which makes Latinos the largest, fastest growing minority in the US.
61, 62
By
COO, approximately 63.4% of Latinos in the US identify as Mexican, 9.5% identify as Puerto
Rican, 3.7% as Cuban, and the remainder are from other CSA countries. By 2060, the proportion
of Latinos in the US is estimated to rise to 28.6% of the total population, or 119 million people.
By state, California has the largest number of Latinos in the US, with 2015 estimates showing a
total of 15.2 million Latinos living in California, 4.9 million of who reside in Los Angeles
County alone.

In the US, overall cancer incidence
63
and mortality rates are lower in Latinos than NLB and
NLW.
64
Despite this, among Latinos, 21.5% of all-cause mortality is due to cancer and cancer is
the leading cause of death, whereas in NLB and NLW, cardiovascular disease is the leading
cause of death, accounting for 23.7% of deaths in both populations.
65
Among Latino men, PC is
the most commonly diagnosed cancer, after non-melanoma skin cancer, with an estimated
13,000 newly diagnosed cases (22% of all cancers) in 2015.
66
Since survival is high, PC is not
the leading cause of death among all cancers, but instead is the fourth most common accounting
for 1,800 (9%) of all cancer-related deaths. Latinos fare better than NLB with respect to disease
prognosis and outcomes,
3
and present with fewer localized PC tumors and more distant tumors
than NLW.
26

14

The available data show that mortality rates from PC are higher among Latinos living in some
CSA nations compared to the US.
67
In general, PC incidence rates for US Latinos are higher than
those of Latinos living outside the US.
58, 68
However, differences in screening strategies and PC
diagnosis make it hard to assess the true burden of disease within the international Latino
community
69
and thus incidence rates may not be comparable.
70
That said, the existing data
suggest that assimilation, environmental exposures, behavioral changes, and other modifiable
risk factors may influence the increasing burden of PC immigrant US Latinos and their
descendants.
68
Though partially attributable to an increase in PSA screening, understanding the
exact cause of the change in incidence patterns proves challenging, partially due to the
heterogeneous nature of Latinos, which descend from generations of admixture from three
ancestral populations: European, African and Indigenous Americans. Moreover, Latinos are
heterogeneous in terms of culture and once they immigrate to the US there is potential further
admixing in subsequent generations after immigration.
71, 72
In studying cancer patterns among
US Latinos, efforts should be made to capture this heterogeneity as best as possible. In this
study, we have chosen COO as an indicator of heterogeneity, in an attempt to capture
subpopulations defined by different admixture patterns, culture, and possibly environmental
exposures.

15
1.4. California Cancer Registry
The California Cancer Registry (CCR) is an internationally recognized component of the
California Department of Public Health Chronic Disease Surveillance and Research Branch,
which monitors and collects state-wide, population-based data from hospitals, clinics, and
medical centers regarding newly diagnosed cancer cases, death, demographics, diagnostics, and
first course of treatment.
73
The goal of the registry is to provide information to researchers and
the general public regarding demographic and geographic associations with cancer risk,
treatment and early detection methods. The CCR contributes significant data to California-based
researchers regarding cancer incidence, mortality and survival among Californian-based
populations irrespective of immigration status. Mandated by law since 1988, the CCR has
required physicians to report cancer-related diagnoses. The purpose of the registry is to monitor
In our study, we utilized the CCR data to determine incidence patterns of PC among all
Californian males, focusing on Latinos stratified by their COO and their incidence relative to
NLB and NLW.

1.5. Research Objectives
The goal of this thesis is to report the frequency of key patient and clinical characteristics of
Californian Latinos with PC diagnosed between 1995 and 2012, in comparison with their NLB
and NLW counterparts, taking into account subpopulations of Latinos defined by COO. By
considering demographic as well as clinical characteristics, such the grade of their tumors at

16
biopsy and treatments they sought, it was our goal to assess whether or not there were disparities
related to race/ethnicity, and among Latinos, across subpopulations defined by COO.

Our objectives were as follows:
Objective 1: Evaluate the differences in the distributions of patient characteristics by
race/ethnicity (NLB, NLW, Latinos), Latinos by COO, and nativity status among primary PC
patients in California, diagnosed from 1995-2012.

Objective 2: Evaluate the differences in the distributions of clinical characteristics by
race/ethnicity (NLB, NLW, Latinos), Latinos by COO, and nativity status among primary PC
patients in California, diagnosed from 1995-2012.

Objective 3: Evaluate differences in treatment patterns among primary PC patients in California,
diagnosed from 1995-2012, by race/ethnicity and Latinos by COO.

Objective 4: Evaluate differences in survival patterns among primary PC patients in California,
diagnosed from 1995-2012, by race/ethnicity and Latinos by COO.

17
2. Methods And Study Design
2.1. Case Identification
We used cancer incidence data collected by the CCR from 380,484 men diagnosed with primary
PC. These data from Californian-based men were identified using SEER site code 28010 and
collected from 1995 through 2012. Primary diagnoses were made using site and histology codes
from the International Classification of Diseases for Oncology, Third Edition (ICD-O-3).
74

Latino status and identification of Latino subpopulations by COO were determined by direct or
indirect identification using the North American Association of Central Cancer Registry
(NAACCR) Hispanic Identification Algorithm (NHIA) version 2.2.1. The NHIA algorithm is a
composite identifier of Latino status derived from variables in NAACCR that classifies patients
according to their Hispanic/Spanish origin (direct identification), family/last name, maiden
name, birthplace/COO, Indian Health Status link, and race from either self-report or observation
from hospital admitting staff (indirect identification), and has a positive predictive value (PPV)
of 94.6%.
75
Racial identity was collected as part of the CCR dataset and coded either based on
self-report or by observation from the hospital admitting staff. We did not subdivide subjects
who identify as multiracial as the option to elect multiple races was only available after the year
2000. In the single-race classification scheme, the most underrepresented racial group of a
multiracial patient was used. We recoded the NHIA-based variable to group Latino
subpopulations by COO. As such, we divided all Latino patients into the following

18
subpopulations: Mexico, Puerto Rico, Cuba, CSA (any country in Central or South America not
specified belonging to this particular region), and not otherwise specified (NOS), i.e. no
information was available for them on birthplace or region of origin.

Of the men with available data, we excluded men who were from Belize, whose PC diagnosis
was coded as “in situ carcinoma of the prostate gland”, whose age was under 40 or over 85 years,
and whose race was Asian or Native American. Because the numbers for the Dominican
Republic were so small, we combined this population’s data into the NOS category. This resulted
in a final study population of 321,433 men, of whom 236,736 (73.7%) were NLW, 33,431
(10.4%) were NLB, and 51,266 (16.0%) were Latino. By COO, among Latinos, there were
13,504 Mexicans (26.3%), 418 Puerto Ricans (0.8%), 671 Cubans (1.3%), 3,551 CSA (6.9%),
and 33,122 NOS (64.6%).

2.2. Study Variables
Age was obtained from the CCR based on the time of diagnosis and categorized as less than 50
years, 50 to 64 years, and greater than 65 years.

SES was defined in accordance to previously established CCR methods
57
, which used data from
the US Census and the American Community Survey depending on year of diagnosis to define
subjects as low/lower-middle, middle, and upper-middle/high SES. Nativity status (US-born vs.

19
Foreign-born) was determined based on birthplace and NHIA. For cases with unknown
birthplace, nativity was estimated using the individual's social security number. Specifically,
among subjects missing nativity status, social security numbers issued prior to the twenty-first
birthday were considered US-natives. Any social security numbers established after the subject’s
twenty-first birthday was considered foreign-born.
76

Insurance status, defined as the primary source of payment to treating hospital, was categorized
as not insured, managed care, Medicare/Medicaid, other and unknown. Vital status was defined
as alive or dead. Marital status was categorized as single or married to better account for
individuals who were either living alone or individuals who were living with someone, which
some studies have shown result in differences in health seeking behaviors, treatment selection
and outcomes.
77, 78

Year of diagnosis was categorized in 6-year intervals. We defined year of diagnosis as 1995
through 2000, 2001 through 2006, and 2007 through 2012. PSA data was divided into two
variables: PSA status and PSA results. Combining data from two CCR variables that measured
whether a PSA test was done generated PSA status variable. PSA status was measured from
1998 through 2012. Our final PSA status variable was defined as yes or no. Among men who
were coded as having had a PSA test performed, we assessed the results of their PSA test based

20
upon the methods for PSA status previously described. The PSA results variable was defined as
negative/normal, positive/elevated, borderline or missing and did not include numerical values
for PSA readings.

The total number of biopsy cores taken and the total number of cores positive for PC were
evaluated categorically. We defined the variables as no cores, 1 – 6 cores, 7 – 12 cores, or more
than 12 cores for analysis. These data were only available for subjects from 2001 through 2012.
Tumor size was determined among patients who had surgery for PC and was defined as the
largest diameter or dimension of the primary tumor in centimeters (cm). We categorized these
data as no tumor, <2 cm, 2 – 4 cm, and 5+ cm. Tumor stage at diagnosis was defined by SEER
Summary Stage as localized, regional, and remote positive. In addition, stage at diagnosis was
classified according to the SEER-modified American Joint Committee of Cancer Collaborative
Stage Data Collection System (Clinical Site version 02.05). PC tumor grading at diagnosis and
among candidates who elected surgery as their treatment option, was evaluated using Gleason
scores. We defined grade groups as follows: not performed, low-grade Gleason 2 – 6,
intermediate-grade Gleason 7, and high-grade Gleason 8 – 10. These data were only available
for subjects from 2001 through 2012.

Treatment status was based on the first-course treatment information available in the CCR data
and defined as no treatment, had treatment, recommended to treatment but did not comply with
treatment recommendations, and unknown. To generate this variable, we combined data from

21
individual variables that captured information regarding treatment with chemotherapy, radiation,
surgery, hormone therapy, combination, or alternative therapies. We also generated a variable
that captured the type of treatment received, including the following categories: hormone,
alternative, radiation, combination, or chemo- therapies, surgery, or unknown. For patients
without any record of having received a treatment we assigned them the category of “possible
active surveillance”. Upon identifying surgical patients among those who had treatment, we also
assessed whether or not a lymphadenectomy was performed during surgery. This variable was
defined as yes vs. no.

Time from diagnosis to treatment was assessed among patients who received treatment.
Defined as the difference between treatment date and diagnosis date, on evaluation, we only
looked at patients with diagnosis dates. For individuals missing data on the day of the month
were, assigned day 15 of the month, if the year and month were recorded in the patient record. If
an individual was missing the month and day but had the year recorded, they were assigned July
1
st
of that year. The final variable was then categorized and defined as follows: 0 – 3 months, 3 –
6 months, 6 – 12 months, > 12 months. Overall survival time was calculated among patients
irrespective of treatment status and included any cause of death, not just PC-caused. Survival
time (in months) was and calculated from the time of treatment until time of death. We did not
restrict estimates by vital status, and person-years among subjects whose vital status was
unknown are considered in the estimates. We considered survival time categorically defined as <
5 years, 5 – 9 years, 10 – 14 years, > 15 + years post-diagnosis.

22

2.3. Data Analysis
We compared frequencies of the variables described above and evaluated them across the three
racial/ethnic groups (Latinos vs. NLW and NLB), among all Latinos by COO (Mexican, Cuban,
Puerto Rican, CSA and NOS), and by nativity status using Pearson’s Chi-Square test. The risk of
death after PC diagnosis associated with race/ethnicity and COO was estimated using Hazard
Ratios (HR) obtained using Cox proportional hazards regression in crude and multivariable
analysis relative to NLW. Multivariable Cox proportional hazards regressions were adjusted for
age, SES, year of diagnosis, PSA, Gleason score at biopsy, and treatment status (treatment:
yes/no). Additional crude and multivariable Cox proportional hazards regression analyses were
conducted stratified by nativity status evaluating the relationship among ethnicity, COO and
survival as previously described. All statistical analyses were performed using Stata 13.1
(StataCorp, College Station, TX, USA). Two-tailed p-values of < 0.05 were considered
statistically significant.

3. Results
3.1. Patient Characteristics
Table 1 describes the key Patient characteristics of PC cases diagnosed in California stratified by
ethnicity (Table 1A) and COO (Table 1B) from 1995 – 2012. Compared to NLB, Latinos had a
statistically significantly lower proportion of cases diagnosed before 65 years (38.1% vs. 51.5%,
p<0.001, Table 1A). We also observed statistically significant differences by nativity status,

23
with a higher proportion of Latinos identifying as being foreign born vs. US-born when
compared to NLB and NLW (57.1% vs. 8.6% vs. 13.1%, p<0.001). Both Latinos (53.5%) and
NLB (50.1%) reported higher proportions of subjects in lower SES groups than NLW (21.8%,
p<0.001). By insurance status, compared to NLB and NLW respectively, Latinos had the highest
proportion of individuals who were uninsured (2.4% vs. 2.1% vs. 0.9%, p<0.001). NLB had the
highest proportion of individuals who reported their marital status as single, widowed, divorced
or separated than Latinos or NLW (39.5% vs. 23.3%, p<0.001). Latinos had the highest
proportion of deceased cases by the end of our study period compared to both NLW and NLB
(71.3% vs. 64.6% vs. 66.2%, p<0.001).

Among Latino subpopulations defined by COO (Table 1B), CSA had a statistically significantly
higher proportion of cases diagnosed before 65 years than any other subpopulation (43.8% vs.
38.2% Mexican vs. 39.0% Puerto Rican vs. 31.9% Cuban vs. 37.6% NOS, p<0.001). By nativity
and insurance status, NOS had the highest proportion of cases who identified as US-born (62.4%
vs. all others < 12.0%, p<0.001) with some form of insurance (98.5% vs. all others < 98.6%,
p<0.001), whereas CSA had the highest proportion of cases identifying as foreign-born (98.9%
vs. all < 98.5%, p<0.001) with the highest proportion of uninsured among all COO (5.7% vs. all
others < 3.6%, p<0.001). We observed statistically significant differences by SES, with
Mexicans having the highest proportion in the lower SES group (64.3% vs. all others <50.1%,
p<0.001). CSA had the highest proportion of deceased cases by the end of our study period

24
(74.7% vs. all others < 73.0%, p<0.001). Puerto Ricans had the highest proportion of individuals
who reported their marital status as single, widowed, divorced or separated than all other
subgroups (32.4% vs. all others < 26.9%, p<0.001).

3.2. Clinical Characteristics and Treatment Received
Table 2 describes the key clinical characteristics of PC cases diagnosed in California stratified
by ethnicity (Table 2A) and COO (Table 2B) from 1995 – 2012.

Clinical Characteristics:
Latinos had the lowest proportion of PC diagnoses from 1995 – 2000 compared to NLW and
NLB (10.9% vs. 15.9% vs. 14.3%, respectively, p<0.001, Table 2A). More NLB received PSA
testing for PC diagnosis than Latinos and NLW (98.8% vs. 98.1% vs. 97.9%, p<0.001), and had
higher proportion of elevated PSA than Latinos and NLW (94.2% vs. 91.8% vs. 89.0%,
p<0.001). Regarding both receipt and PSA results, the proportion of Latinos who received a PSA
test and whose PSA test was elevated fell between that of NLB and NLW. The distribution of the
number of biopsy cores taken at biopsy from 2001 – 2012 was similar among all three groups. Of
the cases that received a biopsy during this timeframe, NLW had the highest proportion of cores
positive for PC on 1 – 6 cores relative to Latinos and NLB (76.4% vs. 74.9% vs. 71.8%,
p<0.001), however, NLB had the highest proportion of cores positive for PC greater than 7 cores
relative to Latinos and NLW (22.7% vs. 18.0% vs. 18.7%, p<0.001). Correspondingly, NLW had
the highest percentage of cases with localized PC relative to Latinos and NLB (84.4% vs. 81.7%

25
vs. 81.9%, p<0.001), but NLB had the highest proportion of cases with lymph node positive PC
and metastases than Latinos or NLW (8.1% vs. 7.7% vs. 5.7%, p<0.001). At the time of biopsy,
Latinos had the highest percentage of low-grade Gleason 2-6 PCs relative to NLW and NLB
(48.3% vs. 47.3% vs. 43.4%, p<0.001) from 2001 – 2012, whereas NLB had the highest
proportion of both intermediate Gleason 7 and high-grade Gleason 8-10 disease (53.3% vs.
Latinos and NLW < 49.4%, p<0.001).

By Latino subpopulation (Table 2B), there were higher proportions of Cubans diagnosed from
1995 – 2000 than any other COO (19.5% vs. < 15.3% for all others, p<0.001). A more
contemporary timeframe (2007 – 2012) shows that NOS had the highest proportion of cases
diagnosed in later years than other groups (59.1% vs. < 54.6% for all others, p<0.001). The
percentage of the cohort who received a PSA test differed marginally among Latinos by COO,
however, among those who received a PSA test, Mexicans were more likely to have an elevated
PSA result than the other subpopulations (92.2% vs. all others <92.0%, p<0.001). The
distribution of the number of biopsy cores taken at biopsy from 2001 – 2012 was similar among
the COO groups. Of the cases that received a biopsy during this time frame, Puerto Ricans had
the highest proportion of cores positive on 1 – 6 cores and localized PC stage relative to all other
COO (~84.0% vs. all others < ~80.0% for both stage and number of cores positive, p<0.001),
however, less than 3% of all Latinos had more than 12 cores positive for PC, irrespective of
COO. Mexicans had the highest proportion of cases with lymph node positive PC and

26
metastases among all COO (10.6% vs. all others < 8.6%, p<0.001). With respect to Gleason
grade on biopsy, both Mexicans and CSA had the highest proportion of high-grade Gleason 8
PC, followed by NOS at 16.1%, relative to the other COO (~17.0% vs. < 16% for all others,
p<0.001). NOS had the highest proportion of low-grade Gleason 2-6 PC at 49.2% (p<0.001).

Treatment Received:
Time from diagnosis to treatment was comparable across the three groups, although NLB had a
slightly higher proportion of men treated before 3 months after diagnosis (94% Latinos vs.
94.1% NLW vs. 96.3% NLB), which was statistically significant. Latinos had a marginally larger
proportion of men who did not receive any treatment compared to NLB or NLW (14.6% vs. all
11.9 – 13.1%, p<0.001). However, among cases who received treatment for PC, there were no
differences in the proportions across groups in receipt of hormonal therapy (all equal to 0.2%),
chemotherapy (all < 0.6%), or combination therapy (all < 3.3%, p<0. 001), though these are not
major treatment categories for PC. NLB had the lowest proportion of men who received surgery
(38.3% vs. all < 41.7%). Latinos were less likely to receive radiation than NLW and NLB
(28.5% vs. 32.2% vs. 31.8%, p<0.001). From 2001 – 2012, only among surgical patients at the
time of surgery or autopsy, Latinos had the highest proportion of low-grade Gleason 2-6 and
high-grade Gleason 8-10 disease (low-grade: 29.6% vs. < 27.0% for NLW and NLB, p<0.001;
high-grade: 8.7% vs. 8.0 – 8.2% for NLW and NLB, p<0.001).

27
Across all COO, there were similar proportions of cases at each demarcation of time from
diagnosis to treatment. Latinos NOS had the highest proportion of cases that reportedly had no
treatment (16.5% vs. 8.4 – 11.6% for all others, p<0.001). Whereas the proportions of cases that
elected to have treatment were similar across all COO (69.2 – 75.9%), the differences among
COO were more visible among cases where treatment was recommended but not received.
Among this subset of patients, 18.5% of Cubans were recommended to receive treatment and
didn’t, which is statistically significantly higher than the other populations whose proportions
ranged from ~12.5% - 17.5% (p<0.001). Among the cases that did receive treatment for PC,
there were minimal differences among the cases reported receipt of hormonal therapy (all <
0.3%), chemotherapy (all < 1.0%), and combination therapy (all < 5%). Relative to other COOs,
Puerto Ricans had the highest percentage of radiation therapy (37.4% vs. all < 29%, p<0.001)
and 47.8% of CSA elected surgery relative vs. less than 45% of all other COOs (p<0.001).
Among surgical patients, Cubans had the highest proportion of low-grade Gleason 2-6 PC on
surgery (42.1% vs. all others < 32%, p<0.001) and CSA had the largest proportion of high-grade
Gleason 8-10 PC (9.3% vs. all others < 8.5%, p<0.001).

3.3. Impact of Nativity on Patient and Clinical Characteristics and Treatment
Received
Table 3 describes the key Patient (Table 3A) and clinical (Table 3B) characteristics of Latino
PC cases diagnosed in California from 1995 – 2012.

28
Relative to US-born Latinos, among key patient characteristics, foreign-born Latinos were
younger at diagnosis, had a high percentage from Mexico (42.2% vs. 7.7% Mexican-descent US-
born Latinos), were more likely to be of lower SES (61.3% vs. 43.8%), uninsured or on
Medicare/Medicaid (49.3% vs. 34.2%), more likely to be deceased upon the conclusion of our
study period (71.0% vs. 68.6%) and more likely to be married or in a domestic partnership
(78.6% vs. 74.2%, all p<0.001, Table 3A).

Among key clinical characteristics, relative to US-born Latinos, foreign-born Latinos were more
likely to be diagnosed contemporarily, specifically from 2007 – 2012 (66.5% vs. US-born <
40%, p<0.001, Table 3B) and were slightly less likely to have a PSA test (98.0% vs. 98.2%,
p<0.001). Among those who had a PSA test, foreign-born Latinos were more likely to have an
elevated PSA (92.0% vs. 91.1%, p<0.001). Foreign-born Latinos were also more likely to have
only 1-6 cores taken on biopsy. There were no differences between US- and foreign-born Latinos
in the number of cores taken that were positive for PC, Gleason score at diagnosis, or PC stage at
diagnosis. Foreign-born cases were more likely to wait more than 12-months between diagnosis
and treatment (3.2% vs. 2.7%, p<0.001).

Foreign-born Latinos were more likely to receive no treatment for PC than US-born Latinos
(14.2% vs. 12.1%, p<0.001). There were no differences among Latinos who elected treatment,
however, among those who were recommended to treatment and didn’t receive treatment, there

29
was a higher percentage of US-born Latinos than foreign-born who did not adhere to treatment
recommendations (15.3% vs. 13.1%, p<0.001). Among those who received treatment, there were
no differences between US- and foreign-born Latinos by hormonal therapy (all 0.2%), other
treatment (10.0-11.0%), chemotherapy (all < 0.6%), surgery (all ~ 42.8%), or combination
therapy (all ~ 3.4 – 3.5%). More US-born Latinos elected radiation therapy among those who
received therapy (30.5% vs. 27.2%, p<0.001), Finally, foreign-born Latinos were more likely to
die before 5 years after diagnosis than US-born Latinos.

3.4. Prostate Cancer Survival Among Californian Latinos
Table 4 describes the crude and multivariable survival analysis (Table 4A) of PC cases
diagnosed in California stratified by nativity status (Table 4B, Table 4C) from 1995 – 2012.

Latinos had the highest proportion of cases surviving less than 5 years after treatment relative to
NLB and NLW (47.4% vs. 45.4% vs. 40.1%, p<0.001). Across all other time points, 5 – 9 years,
10 – 14 years, and more than 15 years, Latinos the lowest proportions relative to the other
race/ethnicities.

By survival time, 48.2% of NOS cases died within 5 years after diagnosis, a significantly higher
proportion than other COO (vs. all others < 47.5%, p<0.001). They also had the lowest
proportion of survival over 15 years (4.0%), whereas Cubans had the highest proportion over 15
years (6.9%, p<0.001).

30

On crude analysis, we found that NLB have a 15% increased risk of death relative to NLW (HR
1.15, 95% CI: 1.12 – 1.17, p<0.001, Table 4A). There was no difference in risk of death between
NLW and Latinos (p=0.32). On multivariable analysis, relative to NLW, NLB were 19% more
likely to die (HR 1.19, 95% CI 1.17 – 1.22), whereas, Latinos now had a 13% decreased risk of
death after controlling for several key Patient and clinical characteristics (HR 0.87, 95% CI 0.85
– 0.89, all p<0.001). By COO, relative to NLW, Mexicans and Cubans had a 12% and 18%
increased risk of death (Mexicans: HR 1.12, 95% CI 1.08 – 1.15; Cubans: HR 1.18, 95% CI 1.05
– 1.33, all p<0.001), respectively, whereas CSA and NOS had a 21% and 4% decreased risk of
death (CSA: HR 0.79, 95% CI 0.74 – 0.85; NOS: HR 0.96, 95% CI 0.94 – 0.98, all p<0.001).
There was no difference in risk of death between Puerto Ricans and NLW on crude or
multivariable analysis. Results for CSA and NOS were little changed by adjustment.

Stratified by nativity status, on crude analysis, US-born NLB and Latinos were 18% and 2%
more likely to die than NLW (NLB HR 1.18, 95% CI 1.16 – 1.21, p<0.001; Latinos HR 1.02,
95% CI 1.00 – 1.05, p=0.04, Table 4B). When compared to foreign-born cases, the results for
NLB were slightly attenuated, dropping down to a 12% increased risk of death relative to
foreign-born NLW (HR 1.12, 95% CI 1.05 – 1.19, p<0.001), whereas foreign-born Latinos now
had a 17% decreased risk of death relative to foreign-born NLW (HR 0.83, 95% CI 0.81 – 0.86,
p<0.001). For US-born Latino subpopulations relative to NLW, only Mexicans had an increased

31
risk of death (HR 1.29, 95% CI 1.19 – 1.39, p<0.001) whereas none of the other COOs were any
different than NLW (all p>0.05). Among foreign-born Latinos, Mexicans (HR 0.91, 95% CI 0.88
– 0.94), CSA (HR 0.66, 95% CI 0.62 – 0.71), and NOS (HR 0.80, 95% CI 0.77 – 0.84, all
p<0.001) all had decreased risk of death relative to NLW. There were no differences between
NLW and Puerto Ricans (p=0.11) or Cubans (p=0.76).

Multivariable adjustment yielded similar results for risk of death for NLB relative to both US-
and foreign-born NLW (US HR 1.19, 95% CI 1.17 – 1.22; Foreign HR 1.21, 95% CI 1.14 – 1.29,
all p<0.001, Table 4C). However, for Latinos, multivariable adjustment showed between 7 –
16% decreased risk of death relative to both US- and foreign-born NLW (US HR 0.93, 95% CI
0.91– 0.96; Foreign HR 0.84, 95% CI 0.81 – 0.87, all p<0.001). Among US-born Latinos, by
subpopulation, there was no association between Puerto Ricans, Cubans and CSA relative to
NLW after multivariable adjustment (all p>0.05). Mexicans had a 13% increased risk of death
relative to NLW (HR 1.13, 95% CI 1.05 – 1.22, p=0.002). For all other US-born Latinos (NOS),
there was an 8% decreased risk of death relative to NLW (HR 0.92, 95% CI 0.89 – 0.94,
p<0.001). The results were similar to the US-born counterparts for foreign-born Puerto Ricans
and Cubans all of which remained unchanged relative to NLW (all p>0.05), however, among
foreign-born Mexicans (HR 0.95, 95% CI 0.91 – 0.98, p=0.006), CSA (HR 0.77, 95% CI 0.72 –
0.83) and for all other Latinos (NOS, HR 0.75, 95% CI 0.72 – 0.78, all p<0.001), there was a

32
decreased risk of death relative to NLW. These results all show significant variability related to
survival among the different Latino subpopulations irrespective of nativity status.

4. Discussion
Accounting for almost 10% of cancer deaths in American men, PC remains the most prevalent
form of cancer in men with an estimated 238,590 new cases diagnosed in 2013
80
, and the highest
burden among NLB. Little research has been done to understand PC cancer patterns among
Latinos. Our study reports differences by Latino subpopulations (by COO), race/ethnicity
(Latino vs. NLB vs. NLW), and nativity status (US-born vs. foreign-born) among Californian
Latinos, and implicates significant disparities within the larger Latino population. More
specifically, we looked at population-level differences in Patient characteristics, PC clinical
features, risk of death, and PC aggressiveness on initial diagnosis. We found that Latinos
presented with fewer new cases with a lower risk of death, and had more cases of low-grade PC
than their NLB and NLW counterparts. We have shown that, relative to NLB and NLW, Latinos
experience different patterns of PC risk and outcomes. We discuss below the broad implications
of these findings and limitations while suggesting future directions for further investigation.

Race, in the US, is both a social and a biological construct that encompasses many different
facets. Socially, race could describe one’s living environment, diet, lifestyle, cultural, and

33
anthropologic choices. Biologically, racial groups may have differing proportion of ancestral
informative markers or single nucleotide polymorphisms
81
, though genetic ancestry is not solely
linked to skin color. Moreover, we used an algorithm to determine Latino ethnicity, when not
provided directly from records, which could also contribute to misclassification. Racial
identification could provide information about behavior and exposure independent of genetic
ancestry. For example, an individual who is genetically 100% Caucasian could identify as Latino
based on COO. This same individual, as a result of their self-reported racial status, could have
specific behaviors that predispose them to cancer. However, a study by Hamilton et al, which
looked at breast cancer outcomes among Latinas and NLB using the NHIA-algorithm showed
that, when comparing the SEER NHIA identification of Latina status with correctly identifying
someone who self-reports as Latina, the sensitivity was 97.7%, specificity 90.7% and the PPV
was 90.1%, which shows that the algorithm, while not 100%, works really well.
82

The role of acculturation and assimilation needs to be considered among immigrant Latinos in
trying to understand cancer patterns. Financial barriers may contribute to disparities in cancer
patterns – i.e. not having the money to travel to receive health care, or taking the time away
from work to go to receive health care – as well as informational ones – for example, not
understanding the importance of screening, medical literature, or knowing what to ask providers
when contact is made. It is of most importance to understand how disease affects these
immigrant populations such that when health care is needed, they have access to it, decreasing

34
the risk of illness caused by poverty.
43
Specific to PC, among Latinos, studies have linked low
SES with lower incidence
57
, but higher rates of late stage disease on diagnosis.
83
Low
educational status has also been linked to risk of not using early detection screening among
Latinos.
84
More studies of SES and incidence are required, which do not lump together Latinos
by groups of convenience, but that help to determine what really makes this population different
than NLW and NLB, and how to assess differences between Latino populations as well.

On it’s own, the relationship between SES and PC is hard to capture in a meaningful way. SES
could mean an assessment of poverty, income, social status, class, or comprise a set of behaviors
which are incumbent to newly immigrated individuals who are more comfortable around their
“own” than assimilating among the majority (i.e. neighborhood enclaves). A 2014 study by
Schupp et al investigated the relationship of nativity on ethnic enclaves and PC survival among
Californian Latinos with invasive PC using data from the CCR.
58
This study included the same
Californian men diagnosed between 1995 and 2008 as in our study, so their analyses overlap
with our study population. This study showed that among Latino PC cases there were more
foreign-born cases, specifically higher stage and grade, and were more likely to live in low SES
neighborhoods relative to US-born. While the authors were not necessarily using enclaves as a
surrogate for SES, they were attempting to capture a different dimension of health disparities that
relates to acculturation. Acculturation, in turn, relates to self-report race and predisposes
individuals to different cancer risks. While it’s clear the correlation among high ethnic enclaves,

35
low SES, and low assimilation is high irrespective of race/ethnic self-identification, their study
sought to capture the role of community in PC diagnosis through neighborhood makeup and
exposures that may influence cancer risk. They found that when adjusting for neighborhood SES,
foreign-born Latinos had a lower risk of death than US-born Latinos. The effect was modified by
the level of ethnic-ness of the enclave, where Latinos living in highly ethnic areas had a slight
survival advantage relative to Latinos living in less ethnic areas.
58
In our data, which includes an
additional four years of data stratified by COO than Schupp et al, we show that SES is still
associated with PC risk and outcome, but is captured differently; these differences are a
testament to the need to understand better the relationship between SES and PC risk and
outcomes.

Here we report that NLB and Latinos, who are relatively similar to NLW, have differences
between them in PSA results, amount of PC found on biopsy (i.e. number of cores positive for
PC) and higher stage and grade on diagnosis. Perhaps this can be attributable to the SES-
race/ethnicity relationship, or perhaps due to genetic susceptibility. More studies really looking
into how SES influences disease biology and aggressiveness are needed.

The main strength of our study is the use of a population-based data from CCR, its large sample
size, overall representativeness and generalizability to multiple populations, its, the inclusion a
wide range of SES, age at diagnosis, and COO for Latinos living in California, which is the state

36
with the largest number of Latinos in the US, as well as the inclusion of a wide range of clinical
variables. Moreover, another strength is the use of a validated algorithm to identify most Latinos
diagnosed with PC in California.
75

Our study also has several limitations. Chief among them is that more than 60% of Latinos had
missing data for COO status. Another limitation is that we did not have data on actual PSA
values, whether or not a digital rectal exam was performed and the outcome, or prostate volume
size. By not having actual PSA values, it is left open to interpretation what “normal”,
“borderline” and “elevated” means with respect to PSA results. Since the cutoff point for PSA
regarding who is recommended treatment as opposed to who is an active surveillance candidate
has changed over time, having these values would have helped us better understand the outcomes
among our patient population with respect to PC risk. Moreover, there was no data on PC family
history or family history of any cancer, both known independent predictors of a positive PC
diagnosis, and of being diagnosed with high-grade PC.
85, 86
More importantly, these features are
part of the decision-making pathway regarding PC detection. Further efforts to improve data
collection by COO will help understand our findings.

Regarding generalizability, our results may not be generalizable to the entire population of US-
Latinos. Even though California has the largest population of Latinos in the US, the ancestral
origin of Californian Latinos might not be representative of the entire US Latino population. We

37
also recognize that we had small numbers of certain Latino populations, e.g. the Dominican
Republic, El Salvador, and Guatemala, which required us to group them together into one COO
group. These countries have unique identifiers, populations and cultures, so it was not ideal to
evaluate them this way. Furthermore, the small numbers of Cubans and Puerto Ricans within the
study made it hard to generalize their results to their respective COO.

This thesis provides evidence that there are interesting disparities between Latinos and NLW
and NLB and within subpopulations of Latinos that deserve further investigation. Whereas
other studies have evaluated these Patient and clinical indices in other cancers among the
Latino population
87, 88
, this is the first study to examine patterns of PC incidence and survival
across Latino subpopulations defined by COO in California.

In summary, this thesis presents evidence that supports the knowledge that health disparities may
play a role in PC risk and aggressiveness through mechanisms, which still need to be explored.
We observed differences in PC patient characteristics, treatment received, and survival among
Latino men in California according to their ethnicity and COO. Furthermore, our findings
highlight a community whose struggles with PC should no longer be ignored: specifically, when
examining differences by Latino COO, Mexicans had a higher risk of mortality than did other
Latino subgroups. This highlights the importance of considering the heterogeneity in this
minority population in understanding the cancer determinants, patterns of care, and survival for

38
PC among Latinos. More strategies for interventions aimed at understanding disease biology in
Latinos, lowering risk, promoting screening, and increasing healthy preventive behaviors can
curb the risk and aggressiveness of PC in Latino men both domestically and abroad.

Table 1A – Patient Characteristics of Prostate Cancer Cases Diagnosed Between 1995 and
2012 in California by Race/Ethnicity (N = 321,433)
Characteristic
Non-Latino White Non-Latino Black All Latinos
N=236,736 N=33,431 N=51,266
N (%)
Age at Diagnosis, years

< 50 4986 (2.1) 1725 (5.2) 1498 (2.9)
50 - 64 83729 (35.4) 15347 (45.9) 18066 (35.2)
≥ 65 148021 (62.5) 16359 (48.9) 31702 (61.8)

Birthplace Data Available 229744 (97.0) 32009 (95.8) 48678 (95.0)
US Born 199682 (86.9) 29242 (91.4) 20871 (42.9)
Foreign Born 30062 (13.1) 2767 (8.6) 27807 (57.1)
Birthplace Data Not Available 6992 (3.0) 1422 (4.3) 2588 (5.1)

Socioeconomic Status

Lower socioeconomic status 51485 (21.8) 16740 (50.1) 27438 (53.5)
Middle socioeconomic status 47974 (20.3) 6884 (20.6) 10401 (20.3)
Upper socioeconomic status 137277 (58.0) 9807 (29.3) 13427 (26.2)

Insurance Status Data Available 221422 (93.5) 31474 (94.1) 46802 (91.3)
Not Insured 1991 (0.9) 663 (2.1) 1113 (2.4)
Managed Care 120894 (54.6) 16780 (53.3) 23448 (50.1)
Medicaid or Medicare 88412 (39.9) 10387 (33.0) 20230 (43.2)
Other 10125 (4.6) 3644 (11.6) 2011 (4.3)
Insurance Status Data Not Available 15314 (6.5) 1957 (5.6) 4464 (8.7)

Vital Status

Alive 79991 (33.8) 11823 (35.4) 14713 (28.7)
Dead 156745 (66.2) 21608 (64.6) 36553 (71.3)

Marital Status Data Available 219442 (92.7) 30558 (92.7) 45160 (88.2)
Single, Widowed, Divorced, Separated 50704 (23.1) 12065 (39.5) 10503 (23.3)
Married/Domestic Partner 168738 (76.9) 18493 (60.5) 34657 (76.7)
Marital Status Data Not Available 17294 (7.3) 2873 (8.6) 6106 (11.9)
Pearson's Chi-Square was used to detect differences between ethnic groups and country of origin for categorical
variables. All p-values were determined to be at the p<0.001 level.

40
Table 1B – Patient Characteristics of Latino Prostate Cancer Cases Diagnosed Between
1995 and 2012 in California by Country of Origin (N = 51,266)

Pearson's Chi-Square was used to detect differences between ethnic groups and country of origin for
categorical variables. All p-values were determined to be at the p<0.001 level.

Characteristics
Mexican Puerto Rican Cuban
Central/
South American Others
N=13,504 N=418 N=671 N=3,551 N=33,122
N (%)
Age at Diagnosis, years

< 50 385 (2.9) 8 (1.9) 11 (1.6) 119 (3.4) 975 (2.9)
50 - 64 4770 (35.3) 155 (37.1) 203 (30.3) 1435 (40.4) 11503 (34.7)
≥ 65 8349 (61.8) 255 (61.0) 457 (68.1) 1997 (56.2) 20644 (62.3)

Birthplace Data Available 13347 (98.8) 414 (99.0) 670 (99.9) 3525 (99.3) 30722 (92.8)
US Born 1608 (12.0) 34 (8.2) 10 (1.5) 39 (1.1) 19180 (62.4)
Foreign Born 11739 (88.0) 380 (91.8) 660 (98.5) 3486 (98.9) 11542 (37.6)
Birthplace Data Not Available 157 (1.2) 4 (1.0) 1 (0.2) 26 (0.7) 2400 (7.3)

Socioeconomic Status

Lower socioeconomic status 8689 (64.3) 186 (44.5) 318 (47.4) 1641 (46.2) 16604 (50.1)
Middle socioeconomic status 2373 (17.6) 97 (23.2) 147 (21.9) 727 (20.5) 7057 (21.3)
Upper socioeconomic status 2442 (18.1) 135 (32.3) 206 (30.7) 1183 (33.3) 9461 (28.6)

Insurance Status Data Available 12555 (93.0) 389 (93.1) 622 (92.7) 3316 (93.4) 29920 (90.3)
Not Insured 450 (3.6) 9 (2.3) 10 (1.6) 190 (5.7) 454 (1.5)
Managed Care 5230 (41.7) 171 (44.0) 285 (45.8) 1539 (46.4) 16223 (54.2)
Medicaid or Medicare 6375 (50.8) 150 (38.6) 320 (51.5) 1457 (43.9) 11928 (39.9)
Other 500 (4.0) 59 (15.2) 7 (1.1) 130 (3.9) 1315 (4.4)
Insurance Status Data Not Available 949 (7.0) 29 (6.9) 29 (6.9) 235 (6.6) 3202 (9.7)

Vital Status

Alive 4437 (32.9) 151 (36.1) 282 (42.0) 897 (25.3) 8946 (27.0)
Dead 9067 (67.1) 267 (63.9) 389 (58.0) 2654 (74.7) 24176 (73.0)

Marital Status Data Available 12676 (93.9) 398 (95.2) 632 (94.2) 3362 (94.7) 28092 (84.8)
Single, Widowed, Divorced, Separated 2755 (21.7) 129 (32.4) 170 (26.9) 793 (23.6) 6656 (23.7)
Married/Domestic Partner 9921 (78.3) 269 (67.6) 462 (73.1) 2569 (76.4) 21436 (76.3)
Marital Status Data Not Available 828 (6.1) 20 (4.8) 39 (5.8) 189 (5.3) 5030 (15.2)

41
Table 2A – Clinical Characteristics of Prostate Cancer Cases Diagnosed Between 1995 and
2012 in California by Race/Ethnicity (N = 321,433)

Characteristic
Non-Latino White Non-Latino Black All Latinos
N=236,736 N=33,431 N=51,266
N (%)
Year of Diagnosis

1995-2000 37609 (15.9) 4769 (14.3) 5562 (10.9)
2001-2006 80326 (33.9) 11116 (33.3) 16386 (32.0)
2007-2012 118801 (50.2) 17546 (52.5) 29318 (57.2)

Had a PSA Test 181923 (76.8) 26330 (78.8) 40320 (78.6)
No 3886 (2.1) 319 (1.2) 753 (1.9)
Yes 178037 (97.9) 26011 (98.8) 39567 (98.1)
PSA Status Not Available 54813 (23.2) 7101 (21.2) 10946 (21.4)

PSA Result Data Available+ 178037 (75.2) 26011 (77.8) 39567 (77.2)
Negative/Normal 14600 (8.2) 1128 (4.3) 2444 (6.2)
Elevated 158384 (89.0) 24492 (94.2) 36315 (91.8)
Borderline 5053 (2.8) 391 (1.5) 808 (2.0)

Number of Biopsy Cores Taken** 155230 (97.7) 22696 (97.9) 37553 (97.1)
No cores taken 131943 (85.0) 18680 (80.6) 31276 (80.9)
1 - 6 cores 2612 (1.7) 531 (2.3) 942 (2.4)
7 - 12 cores 14460 (9.3) 2382 (10.3) 3730 (9.7)
12 + cores 6215 (4.0) 1103 (4.8) 1605 (4.2)
Number of Biopsy Cores Not Available 3645 (2.3) 491 (2.1) 1102 (2.9)

Number of Positive Biopsy Cores&** 25681 (95.3) 4271 (94.8) 6874 (93.2)
All cores negative 67 (0.3) 10 (0.2) 22 (0.3)
1 - 6 cores 20567 (76.4) 3238 (71.8) 5527 (74.9)
7 - 12 cores 4613 (17.1) 920 (20.4) 1185 (16.1)
12 + cores 434 (1.6) 103 (2.3) 140 (1.9)
Number of Positive Biopsy Cores Not
Available 1251 (4.7) 235 (5.2) 505 (6.8)

Pathological Stage at Diagnosis Available 226659 (95.7) 31584 (94.5) 47837 (93.3)
Localized 191380 (84.4) 25872 (81.9) 39057 (81.7)
Regional 22336 (9.9) 3143 (10.0) 5087 (10.6)
Nodes + 3264 (1.4) 481 (1.5) 870 (1.8)
Remote + 9679 (4.3) 2088 (6.6) 2823 (5.9)
Pathological Stage Not Available or Missing 10077 (4.3) 1847 (5.5) 3429 (6.7)

Gleason Score at Biopsy Available** 38306 (24.1) 6059 (26.1) 10933 (28.3)

42

No biopsy performed 1244 (3.3) 203 (3.4) 335 (3.1)
Low Grade (Gleason 2 - 6) 18132 (47.3) 2632 (43.4) 5281 (48.3)
Intermediate Grade (Gleason 7) 13037 (34.0) 2227 (36.8) 3503 (32.0)
High Grade (Gleason 8-10) 5893 (15.4) 997 (16.5) 1794 (16.4)
Gleason Score at Biopsy Not Available 120875 (75.9) 17128 (73.9) 27722 (71.7)
Time from Diagnosis to Treatment (months) 236185 (99.8) 33393 (99.9) 51168 (99.8)
0-3mos 222862 (94.4) 32147 (96.3) 48082 (94.0)
3-6mos 3705 (1.6) 306 (0.9) 882 (1.7)
6-12mos 3722 (1.6) 299 (0.9) 722 (1.4)
12+ mos 5896 (2.5) 641 (1.9) 1482 (2.9)
Time from Diagnosis to Treatment Not
Available 551 (0.2) 38 (0.1) 98 (0.2)
Had Treatment Available 232831 (98.3) 32974 (98.6) 50281 (98.1)
No Treatment 27521 (11.9) 4332 (13.1) 7346 (14.6)
Had Treatment 166594 (71.6) 21775 (66.0) 35869 (71.3)
Recommended but didn't receive 38716 (16.6) 6867 (20.8) 7066 (14.1)
Treatment Status Not Available 3905 (1.7) 457 (1.4) 985 (1.9)
Treatment Type Available 220473 (93.1) 30240 (90.5) 47515 (92.7)
No Treatment 27166 (12.3) 4283 (14.2) 7268 (15.3)
Other Treatment 19027 (8.6) 3663 (12.1) 4969 (10.5)
Hormonal Therapy 483 (0.2) 55 (0.2) 79 (0.2)
Radiation 70969 (32.2) 9608 (31.8) 13525 (28.5)
Chemotherapy 921 (0.4) 177 (0.6) 271 (0.6)
Surgery 95009 (43.1) 11578 (38.3) 19830 (41.7)
Combination Therapy 6896 (3.1) 876 (2.9) 1573 (3.3)
Treatment Type Not Available 16263 (6.9) 3191 (9.5) 3751 (7.3)
Gleason Score at Surgery/Autopsy Available#** 16223 (25.5) 2161 (27.1) 4178 (28.6)
Not performed 1844 (11.4) 184 (8.5) 523 (12.5)
Low Grade (Gleason 2 - 6) 4385 (27.0) 549 (25.4) 1235 (29.6)
Intermediate Grade (Gleason 7) 8663 (53.4) 1256 (58.1) 2057 (49.2)
High Grade (Gleason 8-10) 1331 (8.2) 172 (8.0) 363 (8.7)
Gleason Score at Surgery/Autopsy Not
Available#** 47441 (74.5) 5821 (72.9) 10442 (71.4)

43

**Data available from 2001 - 2012
+ Among men who had a PSA test
& Among men who had a biopsy
# Among men who had surgery

Tumor Size Available# 34715 (36.5) 4390 (37.9) 7312 (36.9)
No Tumor 27 (0.1) 0 (0) 8 (0.1)
< 2cm 23607 (68.0) 3005 (68.5) 5145 (70.4)
2 - 4.99 cm 10806 (31.1) 1346 (30.7) 2101 (28.7)
≥ 5cm 275 (0.8) 39 (0.9) 58 (0.8)
Tumor Size Not Available 60294 (63.5) 7188 (62.1) 12518 (63.1)
Lymphadenectomy Available# 234532 (99.1) 33113 (99.0) 50776 (99.0)
No 176285 (75.2) 26228 (79.2) 38783 (76.4)
Yes 58247 (24.8) 6885 (20.8) 11994 (23.6)
Lymphadenectomy Not Available 2204 (0.9) 318 (1.0) 490 (1.0)

Survival Time, years
< 5 94932 (40.1) 15120 (45.2) 24319 (47.4)
5 - 9.99 76471 (32.3) 10317 (30.9) 15833 (30.9)
10 - 14.99 50214 (21.2) 6240 (18.7) 8957 (17.5)
≥ 15 15119 (6.4) 1754 (5.3) 2157 (4.2)

44
Table 2B – Clinical Characteristics of Prostate Cancer Cases Diagnosed among Latinos
Between 1995 and 2012 in California by Country of Origin (N = 51,266)

Characteristic
Mexican
Puerto
Rican Cuban
Central/South
American Others
N=13,504 N=418 N=671 N=3,551 N=33,122
N (%)
Year of Diagnosis

1995-2000 1675 (12.4) 64 (15.3) 131 (19.5) 398 (11.2) 3294 (10.0)
2001-2006 4520 (33.5) 150 (35.9) 257 (38.3) 1215 (34.2) 10244 (30.9)
2007-2012 7309 (54.1) 204 (48.8) 283 (42.2) 1938 (54.6) 19584 (59.1)

Had a PSA Test 10464 (77.5) 317 (75.8) 478 (71.2) 2855 (80.4) 26206 (79.1)
No 238 (2.3) 7 (2.2) 14 (2.9) 46 (1.6) 448 (1.7)
Yes 10226 (97.7) 310 (97.8) 464 (97.1) 2809 (98.4) 25758 (98.3)
PSA Status Not Available 3040 (22.5) 101 (24.2) 193 (28.8) 696 (19.6) 6916 (20.9)

PSA Result Data Available+ 10226 (75.7) 310 (74.2) 464 (69.2) 2809 (79.1) 25758 (77.7)
Negative/Normal 627 (6.1) 28 (9.0) 27 (5.8) 220 (7.8) 1542 (6.0)
Elevated 9427 (92.2) 279 (90.0) 427 (92.0) 2554 (90.9) 23628 (91.7)
Borderline 172 (1.7) 3 (1.0) 10 (2.2) 35 (1.3) 588 (2.3)
Number of Biopsy Cores
Taken** 9687 (98.0) 277 (97.2) 397 (97.8) 2573 (97.5) 24619 (96.8)
No cores taken 8160 (84.2) 241 (87.0) 362 (91.2) 2193 (85.2) 20320 (82.5)
1 - 6 cores 237 (2.5) 3 (1.1) 8 (2.0) 74 (2.9) 620 (2.5)
7 - 12 cores 930 (9.6) 27 (9.8) 17 (4.3) 212 (8.2) 2544 (10.3)
12 + cores 360 (3.7) 6 (2.2) 10 (2.5) 94 (3.7) 1135 (4.6)
Number of Biopsy Cores Not
Available 202 (2.0) 8 (2.8) 9 (2.2) 65 (2.5) 818 (3.2)
Number of Positive Biopsy
Cores&** 1633 (94.4) 42 (95.4) 42 (95.4) 416 (93.5) 4741 (92.6)
All cores negative 4 (0.2) 0 (0) 0 (0) 2 (0.5) 16 (0.3)
1 - 6 cores 1283 (74.2) 37 (84.1) 35 (79.6) 328 (73.7) 3844 (75.1)
7 - 12 cores 309 (17.9) 5 (11.4) 7 (15.9) 73 (16.4) 791 (15.5)
12 + cores 37 (2.1) 0 (0) 0 (0) 13 (2.9) 90 (1.8)
Number of Positive Biopsy Cores
Not Available 96 (5.6) 2 (4.6) 2 (4.6) 29 (6.5) 376 (7.4)

Pathological Stage at Diagnosis
Available 12642 (93.6) 394 (94.3) 629 (93.7) 3382 (95.2) 30790 (93.0)
Localized 9804 (77.6) 331 (84.0) 511 (81.2) 2706 (80.0) 25705 (83.5)
Regional 1502 (11.9) 37 (9.4) 70 (11.1) 385 (11.4) 3093 (10.1)
Nodes + 288 (2.3) 7 (1.8) 10 (1.6) 94 (2.8) 471 (1.5)
Remote + 1048 (8.3) 19 (4.8) 38 (6.0) 197 (5.8) 1521 (4.9)
Pathological Stage Not Available
or Missing 862 (6.4) 24 (5.7) 42 (6.3) 169 (4.8) 2332 (7.0)

45

Gleason Score at
Biopsy Available** 2679 (27.1) 56 (19.6) 73 (18.0) 655 (24.8) 7470 (29.4)
No biopsy performed 126 (4.7) 2 (3.6) 4 (5.5) 18 (2.8) 205 (2.7)
Low Grade (Gleason
2 - 6) 1233 (46.0) 26 (46.4) 35 (48.0) 311 (47.5) 3676 (49.2)
Intermediate Grade
(Gleason 7) 860 (32.1) 20 (35.7) 25 (34.3) 212 (32.4) 2386 (31.9)
High Grade (Gleason
8-10) 460 (17.2) 8 (14.3) 9 (12.3) 114 (17.4) 1203 (16.1)
Gleason Score at
Biopsy Not Available 7210 (72.9) 229 (80.4) 333 (82.0) 1983 (75.2) 17967 (70.6)
Time from Diagnosis
to Treatment
(months) 13476 (99.8) 418 (100.0) 671 (100.0) 3547 (99.9) 33056 (99.8)
0-3mos 12531 (93.0) 395 (94.5) 633 (94.3) 3350 (94.5) 31173 (94.3)
3-6mos 280 (2.1) 11 (2.6) 14 (2.1) 55 (1.6) 522 (1.6)
6-12mos 218 (1.6) 6 (1.4) 10 (1.5) 49 (1.4) 439 (1.3)
12+ mos 447 (3.3) 6 (1.4) 14 (2.1) 93 (2.6) 922 (2.8)
Time from Diagnosis
to Treatment Not
Available 28 (0.2) 0 (0) 0 (0) 4 (0.1) 66 (0.2)
Had Treatment
Available 13229 (98.0) 405 (96.9) 665 (99.1) 3520 (99.1) 32462 (98.0)
No Treatment 1486 (11.2) 34 (8.4) 74 (11.1) 409 (11.6) 5343 (16.5)
Had Treatment 9960 (75.3) 300 (74.1) 468 (70.4) 2670 (75.9) 22471 (69.2)
Recommended but
didn't receive 1784 (13.5) 71 (17.5) 123 (18.5) 441 (12.5) 4648 (14.3)
Treatment Status Not
Available 275 (2.0) 13 (3.1) 6 (0.9) 31 (0.9) 660 (2.0)
Treatment Type
Available 12597 (93.3) 377 (90.2) 627 (93.4) 3351 (94.4) 30563 (92.3)
No Treatment 1459 (11.6) 32 (8.5) 73 (11.6) 404 (12.1) 5300 (17.3)
Other Treatment 1313 (10.4) 32 (8.5) 81 (12.9) 256 (7.6) 3287 (10.8)
Hormonal Therapy 26 (0.2) 1 (0.3) 1 (0.2) 7 (0.2) 44 (0.1)
Radiation 3532 (28.0) 141 (37.4) 178 (28.4) 928 (27.7) 8746 (28.6)
Chemotherapy 111 (0.9) 2 (0.5) 1 (0.2) 22 (0.7) 135 (0.4)
Surgery 5630 (44.7) 150 (39.8) 276 (44.0) 1601 (47.8) 12173 (39.8)
Combination
Therapy 526 (4.2) 19 (5.0) 17 (2.7) 133 (4.0) 878 (2.9)
Treatment Type Not
Available 907 (6.7) 41 (9.8) 44 (6.6) 200 (5.6) 2559 (7.7)
Gleason Score at
Surgery/Autopsy
Available#** 1132 (27.8) 16 (16.5) 38 (23.8) 311 (26.2) 2681 (29.4)

46

**Data available from 2001 - 2012
+ Among men who had a PSA test
& Among men who had a biopsy
# Among men who had surgery

Not performed 143 (12.6) 2 (12.5) 5 (13.2) 42 (13.5) 331 (12.4)
Low Grade (Gleason
2 - 6) 331 (29.2) 5 (31.3) 16 (42.1) 85 (27.3) 798 (29.8)
Intermediate Grade
(Gleason 7) 555 (49.0) 8 (50.0) 14 (36.8) 155 (49.8) 1325 (49.4)
High Grade
(Gleason 8-10) 103 (9.1) 1 (6.3) 3 (7.9) 29 (9.3) 227 (8.5)
Gleason Score at
Surgery/Autopsy Not
Available#** 2933 (72.2) 81 (83.5) 122 (76.3) 876 (73.8) 6430 (70.6)
Tumor Size
Available# 1938 (34.4) 61 (40.7) 87 (31.5) 583 (36.4) 4643 (38.1)
No Tumor 3 (0.2) 0 (0) 0 (0) 1 (0.2) 4 (0.1)
< 2cm 1394 (71.9) 47 (77.1) 54 (62.1) 403 (63.1) 3247 (69.9)
2 - 4.99 cm 530 (27.4) 13 (21.3) 32 (36.8) 175 (30.0) 1351 (29.1)
≥ 5cm 11 (0.6) 1 (1.6) 1 (1.2) 4 (0.7) 41 (0.9)
Tumor Size Not
Available 3692 (65.6) 89 (59.3) 189 (68.5) 1018 (63.6) 7530 (61.9)
Lymphadenectomy
Available# 13291 (98.4) 408 (97.6) 664 (99.0) 3527 (99.3) 32886 (99.3)
No 9896 (74.5) 318 (77.9) 479 (72.1) 2477 (70.2) 25612 (77.9)
Yes 3395 (25.5) 90 (22.1) 185 (27.9) 1050 (29.8) 7274 (22.1)
Lymphadenectomy
Not Available 213 (1.6) 10 (2.4) 7 (1.0) 24 (0.7) 236 (0.7)
Survival Time, years

< 5 6418 (47.5) 168 (40.2) 245 (36.5) 1523 (42.9) 15965 (48.2)
5 - 9.99 4186 (31.0) 137 (32.8) 242 (36.1) 1132 (31.9) 10136 (30.6)
10 - 14.99 2298 (17.0) 91 (21.8) 138 (20.6) 720 (20.3) 5710 (17.2)
≥ 15 602 (4.5) 22 (5.3) 46 (6.9) 176 (5.0) 1311 (4.0)

47

Table 3A – Patient Characteristics of Latinos by Nativity Status (N = 51,266)

Demographic Characteristics
US Born Foreign Born Unknown
n=20871 n=27807 n=2588

N (%)
Age at Diagnosis, years

< 50 669 (3.2) 742 (2.7) 87 (3.4)
50 - 64 7885 (37.8) 9179 (33.0) 1002 (38.7)
≥ 65 12317 (59.0) 17886 (64.3) 1499 (57.9)

Country/Region of Origin
Mexico 1608 (7.7) 11739 (42.2) 157 (6.1)
Puerto Rico 34 (0.2) 380 (1.4) 4 (0.2)
Cuba 10 (0.1) 660 (2.4) 1 (0.0)
South/Central America 39 (0.2) 3486 (12.5) 26 (1.0)
Not Otherwise Specified 19180 (91.9) 11542 (41.5) 2400 (92.7)
Socioeconomic Status

Lower 9148 (43.8) 17053 (61.3) 1237 (47.8)
Middle 4727 (22.7) 5094 (18.3) 580 (22.4)
Upper 6996 (33.5) 5660 (20.4) 771 (29.8)

Insurance Status 19553 (93.7) 27807 (91.8) 2588 (66.4)
Uninsured 215 (1.0) 812 (2.9) 86 (3.3)
Managed care 11209 (53.7) 11084 (39.9) 1155 (44.6)
Medicaid/Medicare 6925 (33.2) 12905 (46.4) 400 (15.5)
Other 1204 (5.8) 729 (2.6) 78 (3.0)
Insurance Status Not Available 1318 (6.3) 2277 (8.2) 869 (33.6)
Vital Status

Alive 6548 (31.4) 8057 (29.0) 109 (4.2)
Dead 14323 (68.6) 19750 (71.0) 2480 (95.8)
Marital Status 19095 (91.5) 24799 (89.2) 1266 (48.9)
Single, Widowed, Divorced, Separated 4925 (25.8) 5297 (21.4) 281 (22.2)
Married/Domestic Partner 14170 (74.2) 19502 (78.6) 985 (77.8)
Marital Status Data Not Available 1776 (8.5) 3008 (10.8) 1322 (51.1)

48
Table 3B – Clinical Characteristics of Latinos by Nativity Status (N = 51,266)

Clinical Characteristics
US Born Foreign Born Unknown
N=20871 N=27807 N=2588

N (%)
Year of Diagnosis

1995-2000 5575 (26.7) 6716 (24.2) 320 (12.4)
2001-2006 7181 (34.4) 10017 (36.0) 548 (21.2)
2007-2012 8115 (38.9) 11074 (39.8) 1720 (66.5)

Had a PSA Test 16639 (79.7) 21821 (78.5) 1860 (71.9)
No 307 (1.9) 433 (2.0) 13 (0.7)
Yes 16332 (98.2) 21388 (98.0) 1847 (99.3)
PSA Status Not Available 4232 (20.3) 5986 (21.5) 728 (28.1)
PSA Result Data Available+

Negative/Normal 1064 (6.5) 1310 (6.1) 70 (3.8)
Elevated 14879 (91.1) 19682 (92.0) 1754 (95.0)
Borderline 389 (2.4) 396 (1.9) 23 (1.3)

Gleason Score at Biopsy Available** 4035 (26.4) 5633 (26.7) 1265 (55.8)
No biopsy performed 142 (1.5) 195 (3.5) 18 (1.4)
Low Grade (Gleason 2 - 6) 1915 (47.5) 2678 (47.5) 688 (54.4)
Intermediate Grade (Gleason 7) 1310 (32.5) 1816 (32.2) 377 (29.8)
High Grade (Gleason 8-10) 668 (16.6) 944 (16.8) 182 (14.4)
Gleason Score at Biopsy Not Available 11261 (73.6) 15458 (73.3) 1003 (44.2)
Number of Biopsy Cores Taken** 14911 (97.5) 20523 (97.3) 2119 (93.4)
No cores taken 12514 (83.9) 17425 (84.9) 1337 (63.1)
1 - 6 cores 312 (2.1) 535 (2.6) 95 (4.5)
7 - 12 cores 1462 (9.8) 1830 (8.9) 438 (20.7)
12 + cores 623 (4.2) 733 (3.6) 249 (11.8)
Number of Biopsy Cores Not Available 385 (2.5) 568 (2.7) 147 (6.6)

Number of Positive Biopsy Cores&** 2225 (92.8) 2867 (92.5) 680 (87.0)
All cores negative 9 (0.4) 4 (0.1) 1 (0.2)
1 - 6 cores 1766 (79.4) 2219 (77.4) 553 (81.3)
7 - 12 cores 400 (18.0) 581 (20.3) 110 (16.2)
12 + cores 50 (2.3) 63 (2.2) 16 (2.4)
Number of Positive Biopsy Cores Not Available 172 (7.2) 231 (7.5) 102 (13.0)
Pathological Stage at Diagnosis Available 19810 (94.9) 25919 (93.2) 2108 (81.4)
Localized 16293 (82.3) 20887 (80.6) 1877 (89.0)
Regional 2074 (10.5) 2880 (11.1) 133 (6.3)
Nodes + 359 (1.8) 497 (1.9) 14 (0.7)
Remote + 1084 (5.5) 1655 (6.4) 84 (4.0)
Pathological Stage Not Available or Missing 1061 (5.1) 1888 (6.8) 480 (18.6)

49

Time from Diagnosis to Treatment (months) 20829 (99.8) 27754 (99.8) 2585 (99.9)
0-3mos 19597 (94.1) 25967 (93.6) 2518 (97.4)
3-6mos 365 (1.8) 504 (1.8) 13 (0.5)
6-12mos 315 (1.5) 395 (1.4) 12 (0.5)
12+ mos 552 (2.7) 888 (3.2) 42 (1.6)
Time from Diagnosis to Treatment Not
Available 42 (0.2) 53 (0.2) 3 (0.1)

Had Treatment Available 20505 (98.2) 27343 (98.3) 2433 (94.0)
No Treatment 2483 (12.1) 3884 (14.2) 979 (40.2)
Had Treatment 14894 (72.6) 19872 (72.7) 1103 (45.3)
Recommended but didn't receive 3128 (15.3) 3587 (13.1) 351 (14.4)
Treatment Status Not Available 366 (1.8) 464 (1.7) 155 (6.0)

Treatment Type Available 19404 (93.0) 25916 (93.2) 2195 (84.8)
No Treatment 2453 (12.6) 3841 (14.8) 974 (44.4)
Other Treatment 1930 (10.0) 2856 (11.0) 183 (8.3)
Hormonal Therapy 34 (0.2) 41 (0.2) 4 (0.2)
Radiation 5915 (30.5) 7049 (27.2) 561 (25.6)
Chemotherapy 101 (0.5) 161 (0.6) 9 (0.4)
Surgery 8317 (42.9) 11072 (42.7) 441 (20.1)
Combination Therapy 654 (3.4) 896 (3.5) 23 (1.1)
Treatment Type Not Available 1467 (7.0) 1891 (6.8) 393 (15.2)

Gleason Score at Surgery/Autopsy Available#** 41681 (24.3) 10050 (24.8) 5065 (56.7)
Not performed 25248 (60.6) 6435 (64.0) 3950 (78.0)
Low Grade (Gleason 2 - 6) 4777 (11.5) 1146 (11.4) 337 (6.7)
Intermediate Grade (Gleason 7) 9857 (23.7) 2085 (20.8) 680 (13.4)
High Grade (Gleason 8-10) 1799 (4.3) 384 (3.8) 98 (1.9)
Gleason Score at Surgery/Autopsy Not Available 129653 (75.7) 30404 (75.2) 3864 (43.3)
Tumor Size Available# 3185 (38.3) 3950 (35.7) 177 (40.1)
No Tumor 5 (0.2) 3 (0.1) 0 (0.0)
< 2cm 2213 (69.5) 2809 (71.1) 123 (69.5)
2 - 4.99 cm 940 (29.5) 1111 (28.1) 50 (28.3)
≥ 5cm 27 (0.9) 27 (0.7) 4 (2.3)
Tumor Size Not Available 5132 (61.7) 7122 (64.3) 264 (59.9)

Lymphadenectomy Available# 8294 (99.7) 11002 (99.4) 439 (99.5)
No 3541 (42.7) 4947 (45.0) 218 (49.7)
Yes 4753 (57.3) 6055 (55.0) 221 (50.3)
Lymphadenectomy Not Available 23 (0.3) 70 (0.6) 2 (0.5)
Survival Time, years

< 5 9297 (44.6) 13027 (46.9) 1995 (77.1)
5 - 9.99 6543 (31.4) 8899 (32.0) 391 (15.1)

50

**Data available from 2001 - 2012
+ Among men who had a PSA test
& Among men who had a biopsy
# Among men who had surgery

10 - 14.99 4000 (19.2) 4787 (17.2) 170 (6.6)
≥ 15 1031 (4.9) 1094 (3.9) 32 (1.2)

51
Table 4A – Crude and Multivariable Cox Proportional Hazards Models of Ethnicity and
Country of Origin

*Adjusted for age, nativity, socioeconomic status, year of diagnosis, PSA, Gleason Score at biopsy, and
whether or not they had any treatment
**Relative to NLW men
***Among Latinos only, relative to NLW

Characteristics
Hazard Ratio 95% CI P-Value Hazard Ratio 95% CI P-Value
Crude Multivariable*
All combined

Ethnicity**

Non-Latino Black 1.15 1.12 - 1.17 <0.001 1.19 1.17 - 1.22 <0.001
Latino 0.99 0.97 - 1.01 0.32 0.87 0.85 - 0.89 <0.001

Country of Origin***

Mexican 1.12 1.08 - 1.15 <0.001 0.96 0.93 - 1.00 0.03
Puerto Rican 1.04 0.88 - 1.22 0.67 0.98 0.83 - 1.16 0.84
Cuban 1.18 1.05 - 1.33 0.005 1.05 0.93 - 1.18 0.45
Central/South American 0.79 0.74 - 0.85 <0.001 0.78 0.73 - 0.84 <0.001
Not Otherwise Specified 0.96 0.94 - 0.98 <0.001 0.85 0.83 - 0.87 <0.001

52
Table 4B – Crude Cox Proportional Hazards Models of Ethnicity and Country of Origin
Stratified by Nativity Status
Characteristics
US BORN FOREIGN BORN
Hazard Ratio 95% CI P-Value Hazard Ratio 95% CI P-Value
All races

Ethnicity**

Non-Latino Black 1.18 1.16 - 1.21 <0.001 1.12 1.05 - 1.19 <0.001
Latino 1.02 1.00 - 1.05 0.04 0.83 0.81 - 0.86 <0.001

Latinos Only

Country of Origin***

Mexican 1.29 1.19 - 1.39 <0.001 0.91 0.88 - 0.94 <0.001
Puerto Rican 0.96 0.48 - 1.91 0.90 0.87 0.73 - 1.03 0.11
Cuban 1.51 0.57 - 4.02 0.41 0.98 0.87 - 1.11 0.76
Central/South American 0.51 0.21 - 1.24 0.14 0.66 0.62 - 0.71 <0.001
Not Otherwise Specified 1.01 0.98 - 1.03 0.70 0.80 0.77 - 0.84 <0.001

**Relative to NLW men
***Among Latinos only, relative to NLW

53
Table 4C – Multivariable Cox Proportional Hazards Models of Ethnicity and Country of
Origin Stratified by Nativity Status
+

**Relative to NLW men
***Among Latinos only, relative to NLW
+Adjusted for age, socioeconomic status, year of diagnosis, PSA, Gleason at biopsy and whether or not
treatment was obtained
Characteristics
US BORN FOREIGN BORN
Hazard Ratio 95% CI P-Value Hazard Ratio 95% CI P-Value
All races
Ethnicity**

Non-Latino Black 1.19 1.17 - 1.22 <0.001 1.21 1.14 - 1.29 <0.001
Latino 0.93 0.91 - 0.96 <0.001 0.84 0.81 - 0.87 <0.001

Latinos Only

Country of Origin***

Mexican 1.13 1.05 - 1.22 0.002 0.95 0.91 - 0.98 0.006
Puerto Rican 0.83 0.42 - 1.67 0.61 0.98 0.83 - 1.17 0.86
Cuban 1.81 0.68 - 4.81 0.24 1.03 0.91 - 1.17 0.65
Central/South American 0.77 0.32 - 1.85 0.56 0.77 0.72 - 0.83 <0.001
Not Otherwise Specified 0.92 0.89 - 0.94 <0.001 0.75 0.72 - 0.78 <0.001

54
References
1. Torre, L. A., Siegel, R. L., Ward, E. M. et al.: Global Cancer Incidence and Mortality
Rates and Trends--An Update. Cancer Epidemiol Biomarkers Prev, 25: 16, 2016
2. Global Burden of Disease Cancer, C., Fitzmaurice, C., Allen, C. et al.: Global, Regional,
and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With
Disability, and Disability-Adjusted Life-years for 32 Cancer Groups, 1990 to 2015: A
Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol, 3: 524, 2017
3. Siegel, R. L., Miller, K. D., Jemal, A.: Cancer Statistics, 2017. CA Cancer J Clin, 67: 7,
2017
4. Center, M. M., Jemal, A., Lortet-Tieulent, J. et al.: International variation in prostate
cancer incidence and mortality rates. Eur Urol, 61: 1079, 2012
5. Kanavos, P.: The rising burden of cancer in the developing world. Ann Oncol, 17 Suppl
8: viii15, 2006
6. Potosky, A. L., Miller, B. A., Albertsen, P. C. et al.: The role of increasing detection in
the rising incidence of prostate cancer. JAMA, 273: 548, 1995
7. Bray, F., Soerjomataram, I.: The Changing Global Burden of Cancer: Transitions in
Human Development and Implications for Cancer Prevention and Control. In: Cancer:
Disease Control Priorities, Third Edition (Volume 3). Edited by H. Gelband, P. Jha, R.
Sankaranarayanan et al. Washington (DC), 2015
8. Dy, G. W., Gore, J. L., Forouzanfar, M. H. et al.: Global Burden of Urologic Cancers,
1990-2013. Eur Urol, 71: 437, 2017
9. Fidler, M. M., Bray, F., Soerjomataram, I.: The global cancer burden and human
development: A review. Scand J Public Health: 1403494817715400, 2017
10. Khazaei, S., Rezaeian, S., Ayubi, E. et al.: Global Prostate Cancer Incidence and
Mortality Rates According to the Human Development Index. Asian Pac J Cancer Prev,
17: 3793, 2016
11. Angwafo, F. F.: Migration and prostate cancer: an international perspective. J Natl Med
Assoc, 90: S720, 1998
12. Ng, C. J., Teo, C. H., Abdullah, N. et al.: Relationships between cancer pattern, country
income and geographical region in Asia. BMC Cancer, 15: 613, 2015

55
13. Cheng, H. H., Pritchard, C. C., Montgomery, B. et al.: Prostate Cancer Screening in a
New Era of Genetics. Clin Genitourin Cancer, 2017
14. Freedland, S. J., Isaacs, W. B.: Explaining racial differences in prostate cancer in the
United States: sociology or biology? Prostate, 62: 243, 2005
15. Gann, P. H.: Risk factors for prostate cancer. Rev Urol, 4 Suppl 5: S3, 2002
16. Luedeke, M., Rinckleb, A. E., FitzGerald, L. M. et al.: Prostate cancer risk regions at
8q24 and 17q24 are differentially associated with somatic TMPRSS2:ERG fusion status.
Hum Mol Genet, 25: 5490, 2016
17. Marzec, J., Mao, X., Li, M. et al.: A genetic study and meta-analysis of the genetic
predisposition of prostate cancer in a Chinese population. Oncotarget, 7: 21393, 2016
18. Ebot, E. M., Gerke, T., Labbe, D. P. et al.: Gene expression profiling of prostate tissue
identifies chromatin regulation as a potential link between obesity and lethal prostate
cancer. Cancer, 2017
19. Bouvard, V., Loomis, D., Guyton, K. Z. et al.: Carcinogenicity of consumption of red and
processed meat. Lancet Oncol, 16: 1599, 2015
20. Lin, P. H., Aronson, W., Freedland, S. J.: Nutrition, dietary interventions and prostate
cancer: the latest evidence. BMC Med, 13: 3, 2015
21. Hsing, A. W., Tsao, L., Devesa, S. S.: International trends and patterns of prostate cancer
incidence and mortality. Int J Cancer, 85: 60, 2000
22. Shimizu, H., Ross, R. K., Bernstein, L. et al.: Cancers of the prostate and breast among
Japanese and white immigrants in Los Angeles County. Br J Cancer, 63: 963, 1991
23. Kilbourne, A. M., Switzer, G., Hyman, K. et al.: Advancing health disparities research
within the health care system: a conceptual framework. Am J Public Health, 96: 2113,
2006
24. Thomas, S. B., Quinn, S. C., Butler, J. et al.: Toward a fourth generation of disparities
research to achieve health equity. Annu Rev Public Health, 32: 399, 2011
25. Council, N. R.: Eliminating Health Disparities: Measurement and Data Needs.
Washington, DC: The National Academies Press, 2004
26. Ward, E., Jemal, A., Cokkinides, V. et al.: Cancer disparities by race/ethnicity and
socioeconomic status. CA Cancer J Clin, 54: 78, 2004

56
27. Freedland, S. J., Humphreys, E. B., Mangold, L. A. et al.: Risk of prostate cancer-specific
mortality following biochemical recurrence after radical prostatectomy. JAMA, 294: 433,
2005
28. Espaldon, R., Kirby, K. A., Fung, K. Z. et al.: Probability of an abnormal screening
prostate-specific antigen result based on age, race, and prostate-specific antigen
threshold. Urology, 83: 599, 2014
29. Thomas, G., Jacobs, K. B., Yeager, M. et al.: Multiple loci identified in a genome-wide
association study of prostate cancer. Nat Genet, 40: 310, 2008
30. Gaines, A. R., Turner, E. L., Moorman, P. G. et al.: The association between race and
prostate cancer risk on initial biopsy in an equal access, multiethnic cohort. Cancer
Causes Control, 25: 1029, 2014
31. Chinea, F. M., Lyapichev, K., Epstein, J. I. et al.: Understanding PSA and its derivatives
in prediction of tumor volume: Addressing health disparities in prostate cancer risk
stratification. Oncotarget, 8: 20802, 2017
32. Kryvenko, O. N., Lyapichev, K., Chinea, F. M. et al.: Radical Prostatectomy Findings in
White Hispanic/Latino Men With NCCN Very Low-risk Prostate Cancer Detected by
Template Biopsy. Am J Surg Pathol, 40: 1125, 2016
33. Valdovinos, C., Penedo, F. J., Isasi, C. R. et al.: Perceived discrimination and cancer
screening behaviors in US Hispanics: the Hispanic Community Health Study/Study of
Latinos Sociocultural Ancillary Study. Cancer Causes Control, 27: 27, 2016
34. Marlow, N. M., Halpern, M. T., Pavluck, A. L. et al.: Disparities associated with
advanced prostate cancer stage at diagnosis. J Health Care Poor Underserved, 21: 112,
2010
35. Smaje, C.: Not just a Social Construct: Theorising Race and Ethnicity. Sociology, 31:
307, 1997
36. Machery, E. a. F., L: Social Construction and the Concept of Race. Philosophy of
Science, 72: 1208, 2005
37. Marger, M.: Race and ethnic relations: American and global perspectives, 2 ed. Belmont,
California: Wadsworth Publishing Company 1991
38. DeLancey, J. O., Thun, M. J., Jemal, A. et al.: Recent trends in Black-White disparities in
cancer mortality. Cancer Epidemiol Biomarkers Prev, 17: 2908, 2008

57
39. Ozdemir, B. C., Dotto, G. P.: Racial Differences in Cancer Susceptibility and Survival:
More Than the Color of the Skin? Trends Cancer, 3: 181, 2017
40. Rogers, A., Flowers, J., Pencheon, D.: Improving access needs a whole systems
approach. And will be important in averting crises in the millennium winter. BMJ, 319:
866, 1999
41. Gulliford, M., Figueroa-Munoz, J., Morgan, M. et al.: What does 'access to health care'
mean? J Health Serv Res Policy, 7: 186, 2002
42. Greenberg, C. C., Weeks, J. C., Stain, S. C.: Disparities in oncologic surgery. World J
Surg, 32: 522, 2008
43. Peters, D. H., Garg, A., Bloom, G. et al.: Poverty and access to health care in developing
countries. Ann N Y Acad Sci, 1136: 161, 2008
44. Mulley, A. G., Jr.: Developing skills for evidence-based surgery: ensuring that patients
make informed decisions. Surg Clin North Am, 86: 181, 2006
45. Sepucha, K. R., Fowler, F. J., Jr., Mulley, A. G., Jr.: Policy support for patient-centered
care: the need for measurable improvements in decision quality. Health Aff (Millwood),
Suppl Variation: VAR54, 2004
46. Stanford, J. L., Stephenson, R. A., Coyle, L. M., Cerhan, J., Correa, R., Eley, J. W.,
Gilliland, F., Hankey, B., Kolonel, L. N., Kosary, C., Ross, R., Severson, R., West, D.:
Prostate Cancer Trends 1973 - 1995. Edited by S. Program. Bethesda, MD: National
Cancer Institute, 1999
47. Howrey, B. T., Kuo, Y. F., Lin, Y. L. et al.: The impact of PSA screening on prostate
cancer mortality and overdiagnosis of prostate cancer in the United States. J Gerontol A
Biol Sci Med Sci, 68: 56, 2013
48. Force, U. P. S. T.: Screening for Prostate Cancer: Final Recommendation Statement
AHRQ Publication. Annals of Internal Medicine, 12-05160-EF-2, 2012
49. Carter, H. B., Albertsen, P. C., Barry, M. J. et al.: Early detection of prostate cancer:
AUA Guideline. J Urol, 190: 419, 2013
50. Pinsky, P. F., Prorok, P. C., Yu, K. et al.: Extended mortality results for prostate cancer
screening in the PLCO trial with median follow-up of 15 years. Cancer, 123: 592, 2017
51. Society, A. C.: Cancer Facts and Figures 2017. Atlanta, GA: American Cancer Society,
2017

58
52. Siegel, R. L., Sahar, L., Portier, K. M. et al.: Cancer death rates in US congressional
districts. CA Cancer J Clin, 65: 339, 2015
53. Morgan, R. M., Steele, R. J., Nabi, G. et al.: Socioeconomic variation and prostate
specific antigen testing in the community: a United Kingdom based population study. J
Urol, 190: 1207, 2013
54. Moses, K. A., Zhao, Z., Bi, Y. et al.: The impact of sociodemographic factors and PSA
screening among low-income Black and White men: data from the Southern Community
Cohort Study. Prostate Cancer Prostatic Dis, 2017
55. DeNavas-Walt, C. a. P., B. D.: Income and Poverty in the United States: 2014.
Washington, DC: U.S. Census Bureau Current Population Reports, pp. P60, 2015
56. Philips, B. U., Jr., Belasco, E., Markides, K. S. et al.: Socioeconomic deprivation as a
determinant of cancer mortality and the Hispanic paradox in Texas, USA. Int J Equity
Health, 12: 26, 2013
57. Yin, D., Morris, C., Allen, M. et al.: Does socioeconomic disparity in cancer incidence
vary across racial/ethnic groups? Cancer Causes Control, 21: 1721, 2010
58. Schupp, C. W., Press, D. J., Gomez, S. L.: Immigration factors and prostate cancer
survival among Hispanic men in California: does neighborhood matter? Cancer, 120:
1401, 2014
59. Kish, J. K., Yu, M., Percy-Laurry, A. 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: 236, 2014
60. Census, U. S.: Hispanic Origin, 2016
61. Passel, J. S., Cohn, D. V.: U.S. Population Projections: 2005-2050. Social and
Demographic Trends, 2008
62. Bureau, U. S. C.: FFF: Hispanic Heritage Month 2016. Washington, DC, 2016
63. Merrill, R. M., Harris, J. D., Merrill, J. G.: Differences in incidence rates and early
detection of cancer among non-Hispanic and Hispanic Whites in the United States. Ethn
Dis, 23: 349, 2013
64. Pinheiro, P. S., Callahan, K. E., Siegel, R. L. et al.: Cancer Mortality in Hispanic Ethnic
Groups. Cancer Epidemiol Biomarkers Prev, 26: 376, 2017

59
65. Heron, M.: Deaths: Leading Causes for 2014. National Vital Statistics Report, 65, 2016
66. Siegel, R. L., Fedewa, S. A., Miller, K. D. et al.: Cancer statistics for Hispanics/Latinos,
2015. CA Cancer J Clin, 65: 457, 2015
67. Chatenoud, L., Bertuccio, P., Bosetti, C. et al.: Trends in mortality from major cancers in
the Americas: 1980-2010. Ann Oncol, 25: 1843, 2014
68. Gallo, L. C., Penedo, F. J., Espinosa de los Monteros, K. et al.: Resiliency in the face of
disadvantage: do Hispanic cultural characteristics protect health outcomes? J Pers, 77:
1707, 2009
69. Pow-Sang, M., Destefano, V., Astigueta, J. C. et al.: Prostate cancer in Latin America.
Actas Urol Esp, 33: 1057, 2009
70. Tourinho-Barbosa, R. R., Pompeo, A. C., Glina, S.: Prostate cancer in Brazil and Latin
America: epidemiology and screening. Int Braz J Urol, 42: 1081, 2016
71. Price, A. L., Patterson, N., Yu, F. et al.: A genomewide admixture map for Latino
populations. Am J Hum Genet, 80: 1024, 2007
72. Ruiz-Linares, A., Adhikari, K., Acuna-Alonzo, V. et al.: Admixture in Latin America:
geographic structure, phenotypic diversity and self-perception of ancestry based on 7,342
individuals. PLoS Genet, 10: e1004572, 2014
73. Registry, C. C.: About us, 2017
74. Fritz, A., Percy, C., Jack, A., Shanmugaratnam, K., Sobin, L., Parkin, D. M., Whelan, S.:
World Health Organization International Classification of Diseases for Oncology, vol. 3,
2000
75. Group, N. R. a. E. W.: NAACCR Guideline for Enhancing Hispanic/Latino
Identification: Revised NAACCR Hispanic/Latino Identification Algorithm [NHIA
v.2.2.1]. Edited by N. A. A. o. C. C. Registries. Springfield, IL, 2011
76. Gomez, S. L., Quach, T., Horn-Ross, P. L. et al.: Hidden breast cancer disparities in
Asian women: disaggregating incidence rates by ethnicity and migrant status. Am J
Public Health, 100 Suppl 1: S125, 2010
77. O'Malley, D., Dewan, A. A., Ohman-Strickland, P. A. et al.: Determinants of patient
activation in a community sample of breast and prostate cancer survivors.
Psychooncology, 2017

60
78. Tyson, M. D., Andrews, P. E., Etzioni, D. A. et al.: Marital status and prostate cancer
outcomes. Can J Urol, 20: 6702, 2013
79. Varmus, H., Trimble, E. L.: Integrating cancer control into global health. Sci Transl Med,
3: 101cm28, 2011
80. Howlader N, N. A., Krapcho M, et al.: SEER Cancer Statistics Review, 1975-2010,
National Cancer Institute. Bethesda, MD, 2013
81. Shriver, M. D., Parra, E. J., Dios, S. et al.: Skin pigmentation, biogeographical ancestry
and admixture mapping. Hum Genet, 112: 387, 2003
82. Hamilton, A. S., Hofer, T. P., Hawley, S. T. et al.: Latinas and breast cancer outcomes:
population-based sampling, ethnic identity, and acculturation assessment. Cancer
Epidemiol Biomarkers Prev, 18: 2022, 2009
83. Hoffman, R. M., Gilliland, F. D., Eley, J. W. et al.: Racial and ethnic differences in
advanced-stage prostate cancer: the Prostate Cancer Outcomes Study. J Natl Cancer Inst,
93: 388, 2001
84. Reyes-Ortiz, C. A., Camacho, M. E., Amador, L. F. et al.: The impact of education and
literacy levels on cancer screening among older Latin American and Caribbean adults.
Cancer Control, 14: 388, 2007
85. Muller, R. L., Faria, E. F., Carvalhal, G. F. et al.: Association between family history of
prostate cancer and positive biopsies in a Brazilian screening program. World J Urol, 31:
1273, 2013
86. Thomas, J. A., 2nd, Gerber, L., Moreira, D. M. et al.: Prostate cancer risk in men with
prostate and breast cancer family history: results from the REDUCE study (R1). J Intern
Med, 272: 85, 2012
87. Stern, M. C., Fejerman, L., Das, R. et al.: Variability in Cancer Risk and Outcomes
Within US Latinos by National Origin and Genetic Ancestry. Curr Epidemiol Rep, 3:
181, 2016
88. Stern, M. C., Zhang, J., Lee, E. et al.: Disparities in colorectal cancer incidence among
Latino subpopulations in California defined by country of origin. Cancer Causes Control,
27: 147, 2016

Abstract (if available)

Abstract In California, much like the rest of the United States (US), Prostate Cancer (PC) is the most common non-skin cancer among men. However, PC incidence patterns, clinical characteristics, treatment received, and survival are poorly characterized among US Latinos. Latinos comprise a significantly heterogeneous population with unique national identities, demographic characteristics, and ancestry. Since access to care, treatment received and genetic factors may affect prostate risk and survival, the racial admixture among differing ethnic populations may impact PC risk and survival among Latino subpopulations. However little is known about how prostate risk and survival differ among the Latino subpopulations based on their country of origin (COO). As such, we investigated differences in demographic and clinical characteristics related to incidence and survival of 321,433 men diagnosed with PC in California from 1995-2012 accounting for ethnicity and COO using data from the California Cancer Registry (CCR). ❧ Determination of Latino ancestry and COO were identified using the North American Association of Central Cancer Registry (NAACCR) Hispanic Identification Algorithm (NHIA). We grouped Latinos (15.9% of all cases) into the following COO: Mexico (26.3%), Central/South America (CSA, 6.9%), Cuba (1.3%), Puerto Rico (0.8%), and not otherwise specified (NOS, 64.6%). Non-Latino white (NLW, 73.4%) and non-Latino Black (NLB, 10.4%) Americans were used for comparison. Nativity (US-born vs. foreign-born) was defined based on birthplace information available in the CCR records, of which less than 5% among all ethnicities were unknown. Socioeconomic status (SES) was based on the geocoding of the participant's residential address at diagnosis at the census block group level, using a well-established methodology. Frequencies were calculated for patient and clinical characteristics of men with incident prostate cancer by race/ethnicity, Latino subpopulation and nativity status. Patient characteristics included age, SES, insurance status, vital status and marital status. Clinical characteristics included year of diagnosis, Prostate Specific Antigen test (PSA) status and results, number of cores taken on prostate biopsy, of the biopsy cores taken, the number that were positive for PC, pathological stage at diagnosis, Gleason score at diagnosis and after surgery, time from diagnosis to treatment, treatment status, treatment type, tumor size, lymphadenectomy status, and survival time after diagnosis. Pearson’s Chi-square was used to test differences among ethnicities and subgroups. Univariable and multivariable Cox proportional hazards ratios were used to estimate survival in five-year increments by nativity status, ethnicity and Latino subpopulation. ❧ Incidence related: Latinos were older at diagnosis than NLB, with CSA having the greatest proportion of diagnoses under age 50. Latinos were more likely to reside in lower SES areas than NLW and more likely to be uninsured. NLB were more likely to have elevated PSA (94.2%) than NLW (89.0%) and Latinos (91.8%, p<0.001). Latinos and NLB had the highest proportion of non-localized PC relative to NLW (18.3% and 18.1% vs. 15.6%, p<0.001), among them Mexicans having the highest proportion (22.5%). Among all Latinos by nativity status, foreign-born were older, mainly came from Mexico, of low SES and were uninsured. Foreign-born Latinos had a higher percentage of non-localized PC, and more cores positive for PC on biopsy. ❧ Treatment related: Mexicans also had the longest time from diagnosis to treatment relative to other Latino subgroups. Among all men, surgery and radiation were the most used treatments followed by no treatment. Cubans had the lowest receipt of active treatment. Fewer Latinos reported receiving radiation therapy (28.5%) than NLB or NLW (31.8% and 32.2%, p<0.001), among them Puerto Ricans had the highest proportion of radiation treatment whereas CSA had a greater proportion of surgery. Latinos had greater proportion of cases with no treatment reported (15.3%) relative to NLW and NLB. Foreign-born Latinos had more cases treated with radiation. ❧ Survival related: On crude analysis, NLB had greater risk of death (HR 1.15, 95% CI 1.12-1.17, p<0.001) than NLW. There were no differences in survival between Latinos (as a group) and NLW (p=0.32). On multivariable analysis, NLB had similar results (HR 1.19, 95% CI 1.17-1.22, p<0.001), however Latino ethnicity was associated with lower risk of death than NLW after adjusting for various patient and clinical characteristics (HR 0.87, 95% CI 0.85-0.89, p<0.001). Among only Latino subpopulations, relative to NLW, on crude analysis, Mexicans and Cubans had a 12% and 18% increased risk of death (Mexican: HR 1.12, 95% CI 1.08 – 1.15, Cuban: HR 1.18, 95% CI 1.05 – 1.33, all p<0.001) while both CSA and NOS all had 21% and 4%, respectively, decreased risk of death (CSA: 0.79, 95% CI 0.74 – 0.85, NOS: HR 0.96, 95% CI 0.94 – 0.98, all p<0.001). On both crude and multivariable analysis, there were no differences between NLW and Puerto Ricans (p=0.67), however only on multivariable analysis was there no difference between NLW and Cubans (p=0.45). Multivariable adjustment resulted in a 4% decreased risk of death for Mexicans (HR 0.96, 95% CI 0.93 – 1.00, p=0.03), however, for the results little changed after multivariable adjustment for the other COOs. Similar patterns of results were observed when crude and multivariable models were analyzed by nativity status. ❧ We observed differences in PC incidence patterns, treatment received, and survival among Latino men in California according to their ethnicity and COO. This highlights the importance of considering the heterogeneity in this minority population in understanding the cancer determinants, patterns of care, and survival for PC among Latinos.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Disparities in colorectal cancer survival among Latinos in California

PDF

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

PDF

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

PDF

Dietary and supplementary folate intake and prostate cancer risk

PDF

Comparison of Cox regression and machine learning methods for survival analysis of prostate cancer

PDF

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

PDF

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

PDF

Diet quality and pancreatic cancer incidence in the multiethnic cohort

PDF

Genes and environment in prostate cancer risk and prognosis

PDF

Prostate cancer: genetic susceptibility and lifestyle risk factors

PDF

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

PDF

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

PDF

Determinants of delay in care seeking for pelvic floor disorders among Latina women in Los Angeles

PDF

Fish consumption and risk of colorectal cancer

PDF

The evaluation of the long-term effectiveness of zero/low fluoroscopy workflow in ablation procedures for the treatment of paroxysmal and persistent atrial fibrillation

PDF

Ancestral/Ethnic variation in the epidemiology and genetic predisposition of early-onset hematologic cancers

PDF

Risk factors of pelvic floor disorders in the multiethnic cohort study

PDF

Investigating racial and ethnic disparities in patient experiences with care and health services use following colorectal cancer diagnosis among older adults with comorbid chronic conditions

PDF

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

PDF

Functional analyses of androgen receptor structure pertaining to prostate cancer

Asset Metadata

Creator Gaines, Alexis Ruth (author)

Core Title Prostate cancer disparities among Californian Latinos by country of origin: clinical characteristics, incidence, treatment received and survival

School Keck School of Medicine

Degree Master of Science

Degree Program Molecular Epidemiology

Publication Date 09/27/2017

Defense Date 09/08/2017

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

Tag country of origin,disparities,incidence,Latino,OAI-PMH Harvest,prostate cancer,Survival,Treatment

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Stern, Mariana Carla (committee chair)

Creator Email alexisg@usc.edu,argaines07@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c40-438538

Unique identifier UC11264104

Identifier etd-GainesAlex-5785.pdf (filename),usctheses-c40-438538 (legacy record id)

Legacy Identifier etd-GainesAlex-5785.pdf

Dmrecord 438538

Document Type Thesis

Rights Gaines, Alexis Ruth

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