University of Southern California Dissertations and Theses

The study of germ cell tumors and related conditions: an analysis of self-reported data with characterization and comparison of Family History Questionnaires respondents and non-respondents

(USC Thesis Other)

The study of germ cell tumors and related conditions: an analysis of self-reported data with characterization and comparison of Family History Questionnaires respondents and non-respondents

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
THE STUDY OF GERM CELL TUMORS AND RELATED CONDITIONS:
AN ANALYSIS OF SELF-REPORTED DATA WITH CHARACTERIZATION AND
COMPARISON OF FAMILY HISTORY QUESTIONNAIRES RESPONDENTS AND NON-
RESPONDENTS

By
Yunying Zhang
University Of Southern California

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

August 2014

Copyright 2014 Yunying Zhang

Dedication

Dedicated to my family.

ACKNOWLEDGEMENTS

I would like to express appreciation to Dr. Victoria Cortessis for her expertise guidance
throughout my thesis experience and kind support. I would also like to thank my thesis
committee members Dr. Duncan Thomas and Dr. Wendy Mack for their insight and valuable
advice in preparing the thesis. I would like extend my special thanks to my colleagues Anna
Gilmore and Evelyn Sun for their additional encouragement and suggestions. I am also thankful
to my friends Charles Lacson, Petra Zhang, Qiao Meng, Lijun He, and Pamela Minet Lucid for
their unconditional and persistent support.
i

CONTENT
List of Tables .............................................................................................................................. ii
List of Figures .......................................................................................................................... iii
Abstract ...................................................................................................................................... 1
Introduction ................................................................................................................................ 2
Methods...................................................................................................................................... 5
Study Population Enrollment ................................................................................................. 5
Phase 0. ............................................................................................................................... 5
Phase 1. ............................................................................................................................... 6
Phase 2. ............................................................................................................................... 7
Data Analysis ........................................................................................................................ 10
Phase 0. ............................................................................................................................. 10
Phase 1. ............................................................................................................................. 11
Phase 2. ............................................................................................................................. 13
Results ...................................................................................................................................... 13
Phase 0 .................................................................................................................................. 14
Phase 1 .................................................................................................................................. 14
Phase 2 .................................................................................................................................. 23
Discussion ................................................................................................................................ 28
Bibliography ............................................................................................................................ 32
Appendix A .............................................................................................................................. 35
Partial protocol of CGCTRC study .......................................................................................... 35
Purpose of the study ............................................................................................................. 35
Major research question to be addressed in this project ....................................................... 36
Procedure .............................................................................................................................. 37
Appendix B .............................................................................................................................. 39
Family History Questionnaire .................................................................................................. 39
APPENDIX C .......................................................................................................................... 47
Medical History Questionnaire ................................................................................................ 47
ii

List of Tables

Table 1. For Phase 1, the description of time variables as continuous variables for total,
respondents and non-respondents. ........................................................................................... 15
Table 2. For phase 1, the distribution of independent variables as categorical variables for total
respondents (further divided into active respondents and later respondents) and non-respondents
(further divided into active denials and silent denials). ........................................................... 17
Table 3. In phase 1, among the total study population, association between CCR-reported
variables and response to FHQ (respondents vs. non-respondents). ....................................... 20
Table 4. In phase 1, among the respondents, association between CCR-reported variables and
active response (return after first mailing vs. return to subsequent mailings). ........................ 22
Table 5. In phase 1, among non-respondents, association between CCR-reported variables and
active denials (as opposed to silent denials). ........................................................................... 24
Table 6. Distribution of characteristics among respondents from phase 1, further to be invited to
and respond to phase 2 ............................................................................................................. 25
Table 7. Association between bilateral, cryptorchidism, family history of TC or related conditions
(linkage below), CCR-reported variables and response to phase 2 requests: consent, family
history interview, biospecimens sample. ................................................................................. 26

iii

List of Figures
Figure 1 Diagram flow of phase 1 and phase 2. ............................................................ 5

1

Abstract
Background: Although recent decreasing response rates in epidemiologic studies have been
shown to be related to factors such as race/ethnicity and age in several studies, factors affecting
response rates of testicular cancer (TC) patients have barely been studied.
Goal: The goal of this project was to assess factors related to a man’s decision to respond
to screening (complete and return Family History Questionnaire(FHQ) and enrollment
(completing consent process, family history interview (FHI), and biospecimens donation) among
TC patients in the Causes of Germ Cell Tumors and Related Conditions study (CGCTRC).
Method: Case listings were retrieved from the California Cancer Registry; TC cases were asked
to complete screening FHQ, and respondents and non-respondents were further classified
according to their activeness in response and denial, respectively. Selected FHQ respondents
were classified according to response to invitation to fully enroll.
Data analysis: Unconditional logistic regression was used to estimate the crude and adjusted
odds ratios, corresponding 95% confidence intervals, and p-values for association between
individual characteristics and response.
Results: Men with non-Hispanic white ethnicity, bilateral TC, seminoma tumor histology, earlier
birth year, older ages at diagnosis and contact by the study, and less lag year were more likely to
complete screening. Men with family history of TC or related conditions were more likely to
fully enroll.
Conclusion: Time-related variables, demographic factors, and family history were associated
with response. Further research is needed to learn whether addressing psychosocial factors are
also associated with response.
2

Introduction

The importance of research studies can be seen in the evolution of the management of
testicular cancer (TC), a severe and prevalent disease that is treatable with proper medical care.
In 2014, there will be an estimated 8,820 new cases and 380 deaths due to testicular cancer in the
United States (SEER Stat Fact Sheets: Testis Cancer, 2014). While most cancers are diagnosed
later in age, testicular germ cell tumors (TGCTs) are diagnosed most commonly in young men
between the ages of 15 and 35 (Swartout-Corbeil, Frey, & Oberleitner, 2011). Additionally,
incidence rates of testicular cancer have been steadily rising over the past decades, while
significant racial and ethnic differences in incidence cases have also been reported (Huyghe,
Matsuda, & Thonneau, 2003). Although the majority of TC patients survive, young men affected
by TC experience many additional problems, including elevated rates of second primary tumors
(Travis et al, 2005) and numerous severe mortifies: cardiovascular disease (Oeffinger &
Tonorezos, 2011), neurological side effects (Brydøy et al, 2009), and others.
Since 2000, the Causes of Germ Cell Tumors and Related Conditions (CGCTRC), a
study conducted in the Department of Preventive Medicine at the University of Southern
California Keck School of Medicine (see study protocol in Appendix A), has investigated various
risk factors, such as family history of testicular cancer suggesting genetic predisposition or
personal history of cryptorchidism, contributing to an individual’s testicular cancer. In this multi-
phased study, investigators first requested that TC patients respond to a Family History
Questionnaire (FHQ, see Appendix B), a screening questionnaire seeking information about
personal and family histories of testicular cancer, cryptorchidism, and hypospadias. The next
phase involved inviting select FHQ questionnaire respondents to understand and consent to the
study protocol, participate in a Family History Interview (FHI), and provide a biospecimens
3

sample. The CGCTRC study was designed to enable investigators to search for and characterize
inherited variance, associated with TC and the related conditions. Families with multiple
occurrences of disease and related conditions, double primary malignancies (Kuligina, Reiner,
Imyanitov, & Begg, 2011) were particularly desire for these purposes.
Response rate is defined as the proportion of individuals in a base population that
participate in a study. In recent decades, researchers have become alarmed by a trend of
decreasing participation rates in epidemiologic studies (Galea, 2007). In descriptive studies,
when estimating the prevalence of a condition in a population, a high response rate is needed to
assure that prevalence estimations are accurate. For traditional analytic studies, particularly case
control studies, high response is needed to create confidence that the study was not subject to
strong selection bias. For most study types, a high response rate is desired to assure that results
can be generalized to the base population. In the present study, investigators planned to
characterize inherited genetic determinants of disease occurrence, and wished to determine
whether predictive values would pertain to the general base population or only a small subset of
affected people. Low response rates can raise concerns of possible non-participation bias, or
selection bias, when interpreting results for consent-required studies. Macleod & Watt (2008)
conveyed that “the risk-benefit assessment made regarding the use of medical records without
consent should include the benefits of obtaining research evidence based on 100% of the
population and the possibility of inappropriate or insufficient findings if research is confined to
consented populations” (p. 1). If considerable differences exist between participants and non-
participants, the reliability of results of observational studies would be jeopardized (Kho, 2009).
In investigating reasons for participation, it has been reported that individuals’ decisions to
participate in research studies, respond to medical questionnaires, and consent to review of their
4

medical records are associated with their personal characteristics, such as gender, age, ethnicity
group, socioeconomic status, health status, and the sensitive or possibly stigmatizing issue at
hand (Hill, 2013).
Most testis cancer studies report only overall proportion of individuals invited to
participate who are represented in the final analytic set of study participants. We are aware of
only one epidemiologic study of TC, a small scale Australian study, that addressed potential
determinants of response (Smith, King, Butow, & Olver, 2013), the investigators reported no
association between independent variables and TC cases’ preference of completion questionnaire
(postal mail or online questionnaire), but the study population was far too to provide high power
to detect such associations. More frequently, response rates and related factors, such as age,
race/ethnicity, tumor histology type, and age when contacted by the study, among the TGCT
population have rarely been investigated but could reveal considerable information about these
individuals and provide insight for future studies. The response rate of a study is important
because it sheds light on possible selection bias and generalizability of results, but many studies
that do report response rates do not address covariables that may have an effect on responses.
With the hypothesis that older men and men with more TC related conditions are more likely to
participate in the study, the goal of the research reported here was to estimate the association
between response rates and related factors among the TGCT participants in the CGCTRC study.
5

Methods
Study Population Enrollment
This study was conducted in 3 steps: phase 0 (base population selection), phase 1
(screening questionnaire), and phase 2 (attainment of medical history, family history interview,
and biospecimens). This three-stage design was used to enroll a true population-based sample
that contained a greatly enriched population with rare features indicative of genetic risk. A multi-
staged enrollment allowed investigators to identify an adequate number of families with these
attributes using a systematic process to track, sample, and enroll participants in relation to
recorded values of queried characteristics. All materials and protocol for this study were
approved by the University of Southern California Institutional Review Board (IRB) and the
State of California Committee for the Protection of Human Subjects (CPHS).
Phase 0. According to the U.S. Census, the population of California has been steadily
increasing in the past few decades, as has the number of TC diagnoses within the state. In 1980,
California had a population of 23,667,902 and 143 reported TC cases; in 1990, these numbers
rose to 29,760,021 individuals in the population and 644 reported TC cases; and in 2000, there
were to 33,871,648 in the population and 752 reported TC cases (California’s population in
1950, n.d.). The base population for the study consisted of men diagnosed with TC in California
from the inception of member registries in the California Cancer Registry (CCR) of the
California Department of Public Health’s Cancer Surveillance and Research Branch (CSRB).
The CCR provided case listings per year of men diagnosed with germ cell testicular cancer. To
be contacted, the registry needed to believe the man was still living at the time of releasing his
6

identity to the study. From each CCR case listing, the investigators obtained details on
demographics, contact information, and diagnosis history.
Phase 1. Before contacting any individuals, investigators conducted a process of
physician notification. The investigators attempted to contact at least one of the patient’s
physicians of record, informing this individual about the details and purpose of the study and
inquiring whether there was any reason to forgo contacting the patient.
A number of men were not carried over from phase 0 if they had diagnoses of TC that
were not of germ cell origin (e.g. Sertolli or cell tumors or testicular lymphoma). Men who were
determined to be unable to return the questionnaire were also excluded from phase 1 of the study.
Reasons for exclusion were: inability of investigators to find credible current address, having a
mental or physical disability, being deceased at time of mailing, being incarcerated, or having a
language barrier. An addition subset of men though qualifying for phase 1, had not yet been
contacted when the present analysis was initiated. (Figure 1).
The investigators attempted to contact the remaining men (n=14880) by mail, sending
both a cover letter describing the research and with an English or Spanish screening
questionnaire FHQ, which included the research participants’ bill of rights, and a self-addressed
stamped return envelope. Questionnaires had been translated into Spanish by a USC IRB-
approved Spanish translator and had been reviewed and approved by the USC IRB..
If a patient did not respond to his screening questionnaire within approximately 3 months,
he received a follow-up mailing or phone call, or further queries were performed to obtain a new
mailing address by using additional resources, Experian and the California Department of Motor
Vehicles (DMV).
7

After mailing the screening questionnaires FHQ, cases were divided into respondents and
non-respondents, based on whether the investigators received the completed screening
questionnaires. Respondents were dichotomized into Active Respondents (AR), who completed
and returned the FHQ after the first mailing, and Late Respondents (LR), who completed and
returned the FHQ after two or more mailings. Similarly, non-respondents were characterized as
Active Denials (AD), who refused participation, or Silent Denials (SD), from whom
questionnaires were never returned indicating either wrong address or lack of interest in study
participation.
Phase 2. Upon receipt of the completed FHQ, investigators reviewed information
provided to determine eligibility for phase 2. The completed FHQ provided self-reported data on
personal and family histories of TC, cryptorchidism, and/or hypospadias, as well as first-degree
family structure (i.e. parents, siblings, and children). For purposes of subsequent genetic
analysis, the investigators wished to oversample participants who met the criteria of bilateral TC,
cryptorchidism, or family history of TC or related conditions. Therefore, based on FHQ
responses, participants were invited to phase 2 if they met any of the following criteria: personal
history of bilateral TC, cryptorchidism, and hypospadias; family history of TC or related
conditions; or both parents living.
The investigators contacted potential phase 2 study participants via telephone to seek
their further participation and, if they were interested, to initiate an additional process of
informed consent. Participation in phase 2 consisted of completion of the Medical History
Questionnaire (MHQ, see Appendix C), participation in a FHI, providing access to their medical
records, contribution of biospecimens (blood, saliva, or tissue collected at surgery), and
invitation of select family members to also contribute biospecimens. Each FHI was a detailed
8

interview querying family structure, history of all cancers and TC-related conditions among
members of each family, including all first-degree, second-degree (i.e. grandparents,
uncles/aunts, nephews/nieces), and third-degree relatives and any other affected biological
relatives. After the interview was completed, data were entered electronically into pedigree
format using software called Progeny that allowed specification of family members’ biological
relationships. If it was discovered that more than one family member was diagnosed with TC had
belonged to the base population, the investigators would determine which proband case number
to retain (based on which case completed and returned the screening FHQ first) and retire the
case number(s) of the other family member(s), while still maintaining their proband status.
The principal investigator of the study reviewed each family’s pedigree and identified
family members from whom to seek participation. These family members were invited to join the
study by contributing their own biospecimens if they were identified as having a history of TC or
related conditions or if their biospecimens could provide genetic information for single case
families. For example, biospecimens from both parents of a case are instrumental in presenting
information on the genetic basis of their son’s TC.
The preferred form of biospecimens was a sample of blood, collected by use of a mailed
phlebotomy kit. In instances where the participant was unable to provide a blood sample, a saliva
collection kit was mailed, where saliva was deposited directly into an Oragene kit and returned to
the investigators by mail. Tissue collected at surgery was requested directly from the treating
hospital for some cases.
Phase 2 outcomes included: whether the proband completed the process of informed
consent to participate in phase 2 (yes or no); among those who complete the consent process,
whether FHI was completed under the guidance of a trained interviewer (yes or no); and whether
9

Figure 1. Diagram flow of phase 1 and phase 2.
Did not attend to enroll for unable to return the
FHQ (n=2911)
 Refused by MD (n=15)
 Unable to contact due to error (n=1770)
 Never send mail (n=688)
 Mentally disable (n=25)
 Physically disable (n=10)
 Incarcerated (n=39)
 Lingual barrier (n=9)
 Deceased (n=355)
Cases are eventually invited in phase 1
(n=14,880)
Cases from CCR
(n=17,833)
Cases with germ cell tumor
(n=17,791)
Non-respondents
(n=9193)
Respondents
(n= 5687)
Silent Denials
(n=8691)
Active Denials
(n=422)
Active Respondents
(n=3754)
Late Respondents
(n=1933)
Cases are eventually invited in phase 2
(n= 1331)
Consent Returned
(n= 842)
a
FHI Completed
(n= 689)
a
Biospecimens
Received
(n= 722)
a
Ineligible: non-germ cell tumor
histology
(n=42)

*: overlaps occur.
10

biospecimens sample was provided (yes or no). The analyzed data presented here pertain to
probands only.
Data Analysis
Phase 0. All research data were entered and stored in a research database created for the
study using the Common Application Framework Extensible (CAFE, http://cafe.usc.edu/)
platform. This CAFE application was developed by the USC Norris Comprehensive Cancer
Center. Logistic and consistency checks were performed to clean the data and avoid duplicate
enrollment of a single case. Microsoft Access 2007 was linked to CAFE in order to create a
customized database to track every case’s mailing history, contact history, and data reported by
CCR.
Race/ethnicity. Each man’s race was reported by CCR, and his ethnicity was categorized
if his surname appeared on the 1980 Census Spanish Surname list maintained by the Los Angeles
County Cancer Surveillance program. Race/ethnicity was categorized as non-Hispanic white,
non-Hispanic black, non-Hispanic other, and Hispanic. However, if the participant self-identified
differently in MHQ responses, final race/ethnicity delegation was determined by his MHQ
response.
Laterality and histology of testicular germ cell tumors. Testicular germ cell tumor
laterality was classified as unilateral (single primary tumor) or bilateral (two primary tumors).
CCR provided the histological types of the germ cell tumor. The histology types of TGCTs can
be classified into two general categories: seminoma and non-seminoma (Sesterhenn & Davis,
2004). According to The International Germ Cell Consensus Classification (Mead & Stenning,
2007), 90% of all seminomas have good prognoses; therefore, in the event of a mixture
11

composed of both seminoma and non-seminoma elements, the histology was indicated as non-
seminoma/mixture due to the more severe nature of the non-seminoma tumors. Similarly, if a
bilateral patient presented with one tumor as seminoma only and the other as non-
seminoa/mixture, his overall tumor histology was categorized as non-seminoma/mixture.
Histological type was identified based on codes from International Classification of Diseases for
Oncology, Third Revision and further classified into 2 categories: seminoma only (9060,
Dysgerminoma; 9061, Seminoma, NOS; 9062, Seminoma, Anaplastic; 9063, Spermaticytic
Seminoma; 9064, Intratubular Malignant Germ Cell) and non-seminoma/mixture (9065, Germ
Cell Tumor, Nonsemionomatous; 9070, Embryonal Carcinoma, NOS; 9071, Yolk Sac Tumor;
9072, Polyembryoma; 9073, Gonadoblastoma; 9080, Teratoma; 9081, Teratoma carcinoma;
9082, Malignant Teratoma, Undiff; 9083, Malignant Teratoma, Intermediate; 9084, Dermoid
Cyst; 9085, Mixed Germ Cell tumor; 9100, Choriocarcinoma; 9101, Choriocarcinoma combined
w/ other germ cell elements; 9102, Malignant Teratoma, Trophoblastic) according to severity.
Time-related variables. Year of birth and year of diagnosis (first diagnosis for bilateral
patients) were also obtained from the registry. Age at diagnosis (first diagnosis for bilateral
patients) was calculated by subtracting year of birth from year of diagnosis. Age at first mailing
was calculated by subtracting year of birth from year of first mailing, which was recorded for
each mailing. Lag year was calculated by subtracting diagnosis year from year of first mailing;
lag year indicated the time difference between being diagnosed and getting invited to join the
study at phase 1. All time variables were grouped by decade for categorical analysis.
Phase 1. The descriptive parameters of mean, standard deviation, range, minimum, and
maximum for continuous variables were estimated among the base population and separately in
respondents and non-respondents. Overall response rates and strata-specific response rates were
12

calculated as proportions. The frequency and proportion of active respondents and late
respondents among all respondents and of active denials and silent denials among all non-
respondents were tabulated (Table 1). Unconditional logistic regression was used to estimate the
crude odds ratio (in univariable model), adjusted odds ratios (in multivariable model), and
corresponding 95% confidence intervals in order to assess the association between each
independent variable (race/ethnicity, laterality, histology, year of birth, age at diagnosis, age at
first mailing, and lag year) and the binary dependent variable (whether the patient returned the
completed FHQ) (Table 2). Due to high correlation and colinearity among the time-related
variables, analyses estimating the adjusted effect of each time variable were controlled only for
non-time variables. For non-time variables, adjusted effect was controlled for other non-time
variables and year of birth. Consequently, each adjusted model was guaranteed with VIF < 3.
VIF (Variance Inflation Factor) indicates how much the variance increases due to colinearity. As
Menard (1995) suggested, a VIF of 10 or higher “almost certainly indicates a serious colinearity
problem” (p.76); therefore, as the VIF of this model was less than 3, it appeared that colinearity
was not a severe issue.
Further, unconditional logistic regression analysis was also employed to determine the
association of the independent variables on the outcome of active response among respondents
(Table 3) or active denial among non-respondents (Table 4). Among respondents, an active
response was defined as completion and submission of FHQ after the first mailing. Among non-
respondents, an active denial was defined as directly communicating desire to forgo
participation.
13

Phase 2. From the respondents’ FHQ data, the investigators receive information about
bilateral TC, cryptorchidism, and family history of TC or related conditions. If any of these
responses were left blank, the investigators coded them as if the respondent had selected “no”.
Among respondents in phase 1, frequency and proportion of the cases who were invited
to phase 2 were tabulated; among these invitees to phase 2, frequency and proportion of cases
who responded to phase 2 requests were also presented (Table 5). The same unconditional
logistic regression was applied in phase 2. The estimated effect of the independent variables
(race/ethnicity, histology, family history of TC or related conditions, bilateral, cryptorchidism) on
outcome (whether the case responded to phase 2 requests) was evaluated (Table 6). Statistical
analysis was performed by utilizing Stata 12.0 software (Stata Corp., College Station, Texas
77845 USA), and the LOGIT procedure was applied to obtain odds ratios, 95% confidence
intervals, and p-values.
14

Results
Phase 0
The CCR case listings from 2000 to 2008 enumerated 17,833 men with histories of
testicular cancer in California, as shown in Figure 1. Among them, 42 men were excluded due to
non-germ cell tumor histology. The final base population of this study was 17,791 men with
germ cell tumor testicular cancer. An additional 2911 men from the base population were ruled
out because of their inability to return the FHQ: 15 men’s physicians rejected their patients’
participation in the study; 1770 men were unable to be contacted due to error; 688 men were
never sent a mailing from the investigators; 25 men had mental disabilities; 10 men had physical
disabilities; 39 men were incarcerated; 9 men could not answer the questionnaire due to language
barriers; and 355 men were deceased at the time of mailing.
Phase 1
Figure 1 shows a total of 14,880 men were invited to join phase 1, of whom 5687 were
respondents, those who returned their completed FHQ, and 9193 were non-respondents, those
who did not return their FHQ. Among respondents, there were 3754 active respondents, those
who returned their FHQ after the first mailing, and 1933 late respondents, those who returned
their FHQ more than one mailing. Among non-respondents, there were 422 active denials, who
communicated their desire to forgo participation, and 8691 silent denials, whose questionnaires
were never returned to indicate either wrong address or lack of interest to participate.
Time variables in Table 1 show the years of birth, diagnosis ages, ages at first mailing,
and lag years for total phase 1 individuals and separately for respondents and non-respondents.
The mean years of birth are 1958.6 for respondents and 1962.3 for non- respondents. The mean
ages at diagnosis and at first mailing were 35.7 and 44.8, respectively, for respondents
15

Table 1. For Phase 1, the description of time variables as continuous variables for total, respondents and non-respondents.

* Std Dev: standard deviation
Variable Mean±Std Dev Range Mean±Std Dev Range Mean±Std Dev Range
Year of birth 1960.8± 12.3 1905-2006 1958.6± 12.2 1913-2003 1962.3± 12.1 1905-2006
Diagnose age 34.2± 10.4 0-86 35.7± 10.6 0-86 33.3± 10.2 0-86
Age got 1
st
mail 43.4± 11.7 2-100 44.8± 11.7 2-91 42.6± 11.6 2-100
Lag year 9.3± 6.6 1-38 9.2± 6.8 1-38 9.3± 6.5 1-33
Total Among Respondent Among Non-Respondent
16

and 33.3 and 42.6, respectively, for non-respondents. These data show that, on average, the
respondents were slightly older than non-respondents during both diagnosis and first mailing of
phase 1 materials. The mean lag years or time between diagnosis and first mailing, between
respondents and non-respondents appear comparable at 9.2 and 9.3, respectively.
The distribution of response rates based on the categorical variables of race/ethnicity,
histology, laterality, and time-related variables (year of birth, diagnosis age, age at first mailing,
and lag year) in 10-year increments was analyzed as well (Table 2). There was an overall
response rate of 38.22% for the screening questionnaire, with non-Hispanic white men having
the highest response rate at 42.91%, non-Hispanic other men following at 32.99%, Hispanic men
at 24.58%, and non-Hispanic black men with the lowest response rate at 21.79%. In terms of
histology, men diagnosed with non-seminoma/mixture types demonstrated a slightly lower
response rate of 36.85% compared with men diagnosed with seminoma only, who had a response
rate of 39.22%. Unilaterally affected men had a noticeably lower response rate of 37.54%
compared with bilaterally affected men, who responded at a rate of 65.56%. A trend in age was
observed where it appeared that the later the men were born, the lower the response rate: for
those born before 1945, 51.16% responded, while for those born after 1995, only 28.89%
responded. Additionally, men diagnosed at an older age tended to have higher response rates:
only 22.48% of men diagnosed before the age of 10 responded, while 54.17% of men diagnosed
after the age of 60 responded. Higher response rates were also noted when men were older upon
receiving the first mailing from study investigators; for those 20 years or younger when they
received the first mailing, there was a 27.75% response rate, while the response rate for those 70
years or older when receiving the first mailing was 49.34%. Finally, when analyzing lag years
17

Table 2. For phase 1, the distribution of independent variables as categorical variables for total respondents (further divided into active respondents
and later respondents) and non-respondents (further divided into active denials and silent denials).

Total Respondent Non- Response Active Late Active Silence
N=14880 respondent rate(%) respondents respondents denials denials
N (column%) N =5687 N=9193 Overall=38.22 N=3754 N=1933 N=422 N=8771
Race/ethnicity
Non-Hispanic White 10,690 (71.84) 4587 6103 42.91 3057 ( 81.43) 1530 (79.15) 343 (81.28) 5760 (65.67)
Non-Hispanic Black 217 (1.46) 59 158 27.19 43 ( 1.15) 16 (0.83) 5 (1.18) 153 (1.74)
Non-Hispanic Other 767 (5.15) 253 514 32.99 156 ( 4.16) 97 (5.02) 25 (5.92) 489 (5.58)
Hispanic 3206 (21.55) 788 2418 24.58 498 ( 13.27) 290 (15) 49 (11.61) 2369 (27.01)
Histology type
Non-seminoma/mixture 6127 (41.75) 2258 3869 36.85 1432 ( 38.68) 826 (43.27) 146 (35.18) 3723 (43.04)
Seminoma only 8550 (58.25) 3353 5197 39.22 2270 ( 61.32) 1083 (56.73) 269 (64.82) 4928 (56.96)
missing 203 76 127 37.44 52 24 7 120
Laterality
unilateral 14,520 (97.58) 5451 9069 37.54 3605 ( 96.03) 1846 (95.5) 418 (99.05) 8651 (98.63)
bilateral 360 (2.42) 236 124 65.56 149 ( 3.97) 87 (4.5) 4 (0.95) 120 (1.37)
Year of birth
<1945 1251(8.41) 640 611 51.16 455 ( 12.12) 185 (9.57) 71(16.82) 540(6.16)
1945-1954 2906 (19.53) 1309 1597 45.04 904 ( 24.08) 405 (20.95) 128 (30.33) 1469 (16.75)
1955-1964 5106 (34.32) 1991 3115 38.99 1331 ( 35.46) 660 (34.14) 117 (27.73) 2998 (34.19)
1965-1974 3637 (24.45) 1243 2394 34.18 755 ( 20.11) 488 (25.25) 81 (19.19) 2313 (26.38)
1975-1984 1699 (11.42) 438 1261 25.78 260 ( 6.93) 178 (9.21) 20 (4.74) 1241 (14.15)
1985-1994 234 (1.57) 53 181 22.65 41 ( 1.09) 12 (0.62) 5 (1.18) 176 (2.01)
>=1995 45 (0.3) 13 32 28.89 8 ( 0.21) 5 (0.26) 32 (0.36)
missing 2 2 2
Among Total Among Respondents Among Non-Respondents
18

Table 2. Continued.

Diagnosis age
<=10 129 (0.87) 29 100 22.48 988 ( 26.32) 626 (32.38) -- 100 (1.14)
11-20 818 (5.5) 223 595 27.26 20 ( 0.53) 9 (0.47) 13 (3.08) 582 (6.64)
21-30 4756 (31.97) 1614 3142 33.94 148 ( 3.94) 75 (3.88) 90 (21.33) 3052 (34.81)
31-40 5573 (37.46) 2181 3392 39.14 1452 ( 38.68) 730 (37.77) 161 (38.15) 3231 (36.85)
41-50 2652 (17.83) 1168 1484 44.04 807 ( 21.50) 360 (18.62) 101 (23.93) 1383 (15.77)
51-60 709 (4.77) 342 367 48.24 243 ( 6.47) 99 (5.12) 41 (9.72) 326 (3.72)
>60 240 (1.61) 130 110 54.17 96 ( 2.56) 34 (1.76) 16 (3.79) 94 (1.07)
missing 3 3 3
Age at 1st mailing
<=20 209 (1.42) 58 151 27.75 312 ( 8.31) 192 (9.93) 4 (0.95) 147 (1.68)
21-30 1706 (11.47) 504 1202 29.54 46 ( 1.23) 12 (0.62) 29 (6.87) 1173 (13.37)
31-40 4133 (27.78) 1506 2627 36.44 920 ( 24.51) 586 (30.32) 92 (21.8) 2535 (28.9)
41-50 5081 (34.15) 2007 3074 39.5 1330 ( 35.43) 677 (35.02) 130 (30.81) 2944 (33.57)
51-60 2684 (18.04) 1095 1589 40.8 773 ( 20.59) 322 (16.66) 117 (27.73) 1472 (16.78)
61-70 761 (5.11) 367 394 48.23 268 ( 7.14) 99 (5.12) 35 (8.29) 359 (4.09)
>70 304 (2.04) 150 154 49.34 105 ( 2.80) 45 (2.33) 15 (3.55) 139 (1.58)
missing 2 2 2
Lag year
1-5 6130 (41.2) 2424 3706 39.54 1564 ( 41.66) 860 (44.49) 188 ( 44.55) 3515 (40.09)
6-15 5845 (39.28) 2190 3655 37.47 1456 ( 38.79) 734 (37.97) 150 ( 35.55) 3505 (39.97)
16-25 2599 (17.47) 919 1680 35.36 631 ( 16.81) 288 (14.9) 70 ( 16.59) 1610 (18.36)
>26 306 (2.06) 154 152 50.33 103 ( 2.74) 51 (2.64) 14 ( 3.32) 138 (1.57)
missing 3 3 3
19

under 26 years, men with shorter lag years had slightly higher response rates compared with their
longer lag year counterparts: 39.54% response rate for 1 to 5 lag years compared with 35.36%
for 16 to 25 lag years. However, when looking at lag years of 26 or more, the response rate
soared to 50.33%.
The crude and adjusted odds ratios between categorical variables and response to the
FHQ are summarized in Table 3. In the crude univariable model, statistically significant
associations were found between each of the following variables and response: race/ethnicity
(p<0.0001), laterality (p=0.0057), histology (p=0.0004), year of birth (p<0.0001), age at
diagnosis (p<0.0001), age at first mailing (p<0.0066), and lag year (p<0.0001). Non-Hispanic
black, non-Hispanic other, and Hispanic men appeared to be only 0.50, 0.66, and 0.43,
respectively, times as likely to respond to the FHQ as non-Hispanic white men. Men with
bilateral TC were 3.17 times as likely to respond as men with unilateral TC. Men who were born
before 1945 had the highest response rates, after which a decreasing response rate was observed
with increasingly later years of birth. Higher likelihood of response was observed among men
who were diagnosed and who received their first mailing at older ages. Men with 6 to 25 lag
years were less likely to respond compared with men with less than 6 lag years; however, when
the lag year was greater than 25, men were 1.55 (95% CI: 1.23-1.95, p<0.001) times as likely to
respond as men with less than 6 lag years. All of the above associations remain significant after
adjustment. The only association that was no longer statistically significant after adjustment was
between histology type and FHQ response. Men with less severe histology (seminoma only)
were 1.11 (95% CI: 1.03-1.18, p=0.004) times as likely to respond as men with more severe
histology (non-seminoma/mixture); after adjusting for histology type, this association was
reversed (OR=0.93, 95% CI: 0.87-1.00), although this finding was not statistically significant
20

Table 3. In phase 1, among the total study population, association between CCR-reported variables and
response to FHQ (respondents vs. non-respondents).

*Non-time variables were adjusted for two other non-time variables and year of birth.
E.g.: association between race/ethnicity and response was adjusted for histology, laterality, and
year of birth.
*Time variables were adjusted for non-time variables.
E.g.: association between year of birth and response was adjusted for race/ethnicity, histology,
and laterality.
OR and 95% CI P>|z| sample size Ovelall P-vlaue OR and 95% CI P>|z|
Race/ethnicity
Non-Hispanic white 1 (ref) 1 (ref)
Non-Hispanic black 0.50 (0.37- 0.67) <0.001 0.49 ( 0.36- 0.67) <0.001
Non-Hispanic other 0.66 (0.56- 0.77) <0.001 14800 <0.0001 0.68 ( 0.58- 0.79) <0.001
Hispanic 0.43 (0.40- 0.47) <0.001 0.50 ( 0.46- 0.55) <0.001
Laterality
Unilateral 1 (ref) 1 (ref)
Bilateral 3.17 (2.55- 3.95) <0.001 14880 0.0057 3.14 ( 2.50- 3.93) <0.001
Histology
Non-seminoma/mixture 1 (ref) 1 (ref)
Seminoma only 1.11 (1.03- 1.18) 0.004 14677 0.0004 0.93 ( 0.87- 1.00) 0.061
Year of birth
<1945 1 (ref) 1 (ref)
1945-1954 0.78 (0.69- 0.89) <0.001 0.74 ( 0.65- 0.85) <0.001
1955-1964 0.61 (0.54- 0.69) <0.001 0.59 ( 0.52- 0.67) <0.001
1965-1974 0.50 (0.44- 0.56) <0.001 14878 <0.0001 0.51 ( 0.45- 0.59) <0.001
1975-1984 0.33 (0.28- 0.39) <0.001 0.37 ( 0.32- 0.44) <0.001
1985-1994 0.28 (0.20- 0.39) <0.001 0.35 ( 0.25- 0.49) <0.001
>=1995 0.39 (0.20- 0.75) 0.005 0.52 ( 0.27- 1.00) 0.051
Diagnosis age
<=10 0.56 (0.37- 0.86) 0.007 0.74 ( 0.48- 1.13) 0.161
11-20 0.73 (0.62- 0.86) <0.001 0.77 ( 0.65- 0.91) 0.003
21-30 1 (ref) 1 (ref)
31-40 1.25 (1.15- 1.36) <0.001 14877 <0.0001 1.23 ( 1.13- 1.34) <0.001
41-50 1.53 (1.39- 1.69) <0.001 1.44 ( 1.30- 1.60) <0.001
51-60 1.81 (1.55- 2.13) <0.001 1.75 ( 1.48- 2.07) <0.001
>60 2.30 (1.77- 2.99) <0.001 2.16 (1.66-2.82) <0.001
Age at 1st mailing
<=20 0.90 (0.66-1.24 ) 0.536 1.06 (0.76-1.46 ) 0.744
21-30 1 (ref) 1 (ref)
31-40 1.37 (1.21-1.54 ) <0.001 14879 0.0066 1.21 (1.07-1.37 ) 0.003
41-50 1.56 (1.38-1.75 ) <0.001 1.27 (1.12-1.44 ) <0.001
51-60 1.64 (1.44-1.87 ) <0.001 1.30 (1.13-1.50 ) <0.001
61-70 2.22 (1.86-2.65 ) <0.001 1.81 (1.50-2.18 ) <0.001
>70 2.32 (1.81-2.98 ) <0.001 1.91 (1.47-2.49 ) <0.001
Lag year
1-5 1 (ref) 1 (ref)
6-15 0.92 ( 0.85- 0.99) 0.019 0.82 ( 0.76- 0.89) <0.001
16-25 0.84 ( 0.76- 0.92) <0.001 14877 <0.0001 0.74 ( 0.67- 0.81) <0.001
>25 1.55 ( 1.23- 1.95) <0.001 1.68 ( 1.26- 2.23) <0.001
Crude Partially Adjusted *
21

(p=0.061).
The associations between active responses (dependent variable) and race/ethnicity, laterality,
histology, year of birth, diagnosis age, age at first mailing, and lag year (independent variables)
are presented in Table 4. No statistically significant associations between active responses and
race/ethnicity (p=0.0735), histology (p=0.3439), and lag year (p=0.1346) were detected.
However, significant associations were observed between active responses and laterality
(p=0.0009), year of birth (p<0.0001), diagnosis age (p<0.0001), and age at first mailing
(p<0.0001). The following associations remained significant after adjustment: men with bilateral
TC were 1.21 (95% CI: 1.08-1.35, p=0.001) times as likely to actively respond as men with
unilateral TC; men born from 1965 to 1984 were less likely to actively respond compared with
men born before 1945; men diagnosed after age 30 were also more likely to be active
respondents compared with men diagnosed from ages 21 through 30; men who were sent their
first mailings at age 20 or under and from ages 51 to 70 were more likely to be active in
responding compared with men who were sent their first mailings from ages 21 through 30.
The associations between active denials and the independent variables are shown in Table 5.
Statistically significant associations were observed between active denials and race/ethnicity
(p<0.0001), year of birth (p<0.0001), diagnosis age (p<0.0001), age at first mailing (p<0.0001),
and lag year (p<0.0001). The following associations maintained their statistical significance after
adjustment: Hispanics were less likely to be active denials compared with non-Hispanic whites;
men born after 1945 were less likely to be active denials when compared with men born in 1945
or earlier; men diagnosed after age 30 were also more likely to actively deny when compared
with men diagnosed at ages 21 through 30; men who were sent their first mailing after age 50
were more likely to actively deny compared with men who were sent their first mailing at ages
22

Table 4. In phase 1, among the respondents, association between CCR-reported variables and active
response (return after first mailing vs. return to subsequent mailings).

*Non-time variables were adjusted for two other non-time variables and year of birth.
E.g.: association between race/ethnicity and active response was adjusted for histology, laterality,
and year of birth.

*Time variables were adjusted for non-time variables.
E.g.: association between year of birth and active response was adjusted for race/ethnicity,
histology, and laterality.
OR and 95% CI P>|z| sample size Ovelall P-vlaue OR and 95% CI P>|z|
Race/ethnicity
Non-Hispanic white 1 (ref) 1 (ref)
Non-Hispanic black 1.35 ( 0.76- 2.40) 0.314 1.25 (0.7-2.24) 0.446
Non-Hispanic other 0.80 ( 0.62- 1.04) 0.103 5687 0.0735 0.82 (0.63-1.07) 0.138
Hispanic 0.86 ( 0.73- 1.01) 0.059 0.93 (0.79-1.1) 0.400
Laterality
Unilateral 1 (ref) 1 (ref)
Bilateral 1.21 ( 1.08- 1.35) 0.001 5687 0.0009 1.13 (1.01-1.27) 0.041
Histology
Non-seminoma/mixture 1 (ref) 1 (ref)
Seminoma only 0.88 ( 0.67- 1.15) 0.341 5611 0.3439 0.86 (0.66-1.13) 0.286
Year of birth
<1945 1 (ref) 1 (ref)
1945-1954 0.91 ( 0.74- 1.12) 0.359 0.93 (0.75-1.15) 0.482
1955-1964 0.82 ( 0.67- 1.00) 0.046 0.84 (0.69-1.03) 0.092
1965-1974 0.63 ( 0.51- 0.77) <0.001 5687 <0.0001 0.66 (0.53-0.82) <0.001
1975-1984 0.59 ( 0.46- 0.77) <0.001 0.64 (0.49-0.83) 0.001
1985-1994 1.39 ( 0.71- 2.70) 0.333 1.55 (0.79-3.03) 0.204
>=1995 0.65 ( 0.21- 2.01) 0.456 0.76 (0.24-2.36) 0.631
Diagnosis age
<=10 1.41 ( 0.64- 3.11) 0.398 1.59 (0.71-3.53) 0.257
11-20 1.25 ( 0.93- 1.68) 0.138 1.31 (0.97-1.77) 0.074
21-30 1 (ref) 1 (ref)
31-40 1.26 ( 1.10- 1.44) 0.001 5687 <0.0001 1.22 (1.06-1.41) 0.004
41-50 1.42 ( 1.21- 1.67) <0.001 1.36 (1.15-1.61) <0.001
51-60 1.56 ( 1.21- 2.01) 0.001 1.5 (1.16-1.95) 0.002
>60 1.79 ( 1.19- 2.68) 0.005 1.68 (1.11-2.53) 0.013
Age at 1st mailing
<=20 2.36 ( 1.22- 4.57) 0.011 2.5 (1.29-4.84) 0.007
21-30 1 (ref) 1 (ref)
31-40 0.97 ( 0.79- 1.19) 0.745 5687 <0.0001 0.93 (0.75-1.15) 0.512
41-50 1.21 ( 0.99- 1.48) 0.066 1.14 (0.92-1.4) 0.228
51-60 1.48 ( 1.18- 1.84) 0.001 1.37 (1.08-1.73) 0.008
61-70 1.67 ( 1.24- 2.23) 0.001 1.53 (1.13-2.08) 0.006
>70 1.44 ( 0.97- 2.13) 0.071 1.34 (0.89-2.02) 0.166
Lag year
1-5 1 (ref) 1 (ref)
6-15 1.09 (0.97-1.23) 0.162 1.09 (0.96-1.23) 0.192
16-25 1.2 (1.02-1.42) 0.025 5687 0.1346 1.2 (1.01-1.41) 0.034
>25 1.11 (0.79-1.57) 0.552 1.24 (0.83-1.85) 0.302
Partially Adjusted * Crude
23

21 through 30; and men with 6 to 15 lag years were less likely to be active denials when
compared with those with 1 to 5 lag years, while men with more than 25 lag years were more
likely to be active denials.
Phase 2
As shown in Table 6, individuals with bilateral TC, cryptorchidism, or family history of
TC or related conditions were oversampled for purposes of subsequent genetic analysis. Among
those who completed phase 1 by submitting completed FHQ, 87.71% of men with bilateral TC,
48.94% of men with cryptorchidism, and 90.15 % of men with family history of TC or related
conditions were invited to phase 2. The overall response rates for completion of consent, FHI,
and biospecimens samples were 63.26%, 52.07%, 54.55% respectively.
The associations between responses to phase 2 requests (dependent variables) and history
of bilateral TC, history of cryptorchidism, family history of TC or related conditions,
race/ethnicity, histology, and year of birth (independent variables) are summarized in Table 7.
Men with family histories of TC or related conditions were statistically significantly more likely
to respond to phase 2 requests (p<0.001) compared to men without these family histories. Men
who themselves had bilateral TC or cryptorchidism were less as likely to respond to consent
request as men who had neither bilateral TC nor cryptorchidism, but this association was no
longer significant after adjustment (p>0.05). Men with bilateral TC or cryptorchidism were less
likely to respond to FHI request in unadjusted model, but more likely to respond after
adjustment; however the associations were not significant before or after adjustment. Men with
bilateral TC or cryptorchidism were less likely to respond to biospecimens request, but this
association was also not significant. No significant association was observed in race/ethnicity
after adjustment (p>0.05). The significant association between histology and phase 2 response
24

Table 5. In phase 1, among non-respondents, association between CCR-reported variables and active
denials (as opposed to silent denials).

*Non-time variables were adjusted for two other non-time variables and year of birth.
E.g.: association between race/ethnicity and active denial was adjusted for histology, laterality,
and year of birth.

*Time variables were adjusted for non-time variables.
E.g.: association between year of birth and active denial was adjusted for race/ethnicity,
histology, and laterality.
OR and 95% CI P>|z| sample size Ovelall P-vlaue OR and 95% CI P>|z|
Race/ethnicity
Non-Hispanic white 1 (ref) 1 (ref)
Non-Hispanic black 0.55 (0.22-1.35) 0.184 0.57 (0.23-1.41) 0.226
Non-Hispanic other 0.86 (0.57-1.30) 0.472 9193 <0.0001 0.88 (0.57-1.35) 0.552
Hispanic 0.35 (0.26-0.47) <0.001 0.44 (0.32-0.61) <0.001
Laterality
Unilateral 1 (ref) 1 (ref)
Bilateral 0.69 (0.25-1.88) 0.465 9193 0.4412 0.73 (0.27-2.00) 0.538
Histology
Non-seminoma/mixture 1 (ref) 1 (ref)
Seminoma only 1.39 (1.13-1.71) 0.002 9066 0.0014 1.04 (0.84-1.29) 0.694
Year of birth
<1945 1 (ref) 1 (ref)
1945-1954 0.66 (0.49-0.90) 0.008 0.61 (0.45-0.84) 0.002
1955-1964 0.30 (0.22-0.41) <0.001 0.29 (0.21-0.40) <0.001
1965-1974 0.27 (0.19-0.37) <0.001 9159 <0.0001 0.28 (0.20-0.39) <0.001
1975-1984 0.12 (0.07-0.21) <0.001 0.15 (0.09-0.25) <0.001
1985-1994 0.22 (0.09-0.55) <0.001 0.24 (0.08-0.67) 0.007
>=1995 -- --
Diagnosis age
<=10 -- --
11-20 0.76 (0.42-1.36) 0.353 0.76 (0.41-1.40) 0.370
21-30 1 (ref) 1 (ref)
31-40 1.69 (1.30-2.20) <0.001 9090 <0.0001 1.56 (1.19-2.05) 0.001
41-50 2.48 (1.85-3.32) <0.001 2.22 (1.63-3.01) <0.001
51-60 4.26 (2.89-6.30) <0.001 3.65 (2.43-5.47) <0.001
>60 5.77 (3.25-10.25) <0.001 5.11 (2.86-9.16) <0.001
Age at 1st mailing
<=20 1.10 (0.38-3.18) 0.859 0.95 (0.28-3.16) 0.930
21-30 1 (ref) 1 (ref)
31-40 1.47 (0.96-2.24) 0.074 9191 <0.0001 1.23 (0.80-1.89) 0.351
41-50 1.79 (1.19-2.69) 0.005 1.36 (0.89-2.07) 0.154
51-60 3.21 (2.12-4.88) <0.001 2.33 (1.51-3.58) <0.001
61-70 3.94 (2.36-6.58) <0.001 2.90 (1.71-4.92) <0.001
>70 4.36 (2.27-8.39) <0.001 3.52 (1.81-6.85) <0.001
Lag year
1-5 1 (ref) 1 (ref)
6-15 0.80 (0.64-1.00) 0.046 0.71 (0.57-0.89) 0.003
16-25 0.81 (0.61-1.08) 0.148 9190 0.0148 0.74 (0.56-0.98) 0.036
>25 1.90 (1.07-3.35) 0.025 2.50 (1.30-4.82) 0.006
Partially adjusted* Crude
25

Table 6. Distribution of characteristics among respondents from phase 1, further to be invited to and
respond to phase 2.

*: based on self-reported FHQ; linkage refers to family history of TC or related conditions
a: denominator is the number of men who completed the FHQ in phase 1
b: denominator is the number of men who were invited to phase 2
Phase 1 Completed
N N %
a
N %
b
N %
b
N %
b
Bilateral*
No 5451 1124 20.62 723 64.32 589 52.40 619 55.07
Yes 236 207 87.71 119 57.49 104 50.24 107 51.69
Cryptorchidism*
No 5074 1031 20.32 669 64.89 545 52.86 576 55.87
Yes 613 300 48.94 173 57.67 148 49.33 150 50.00
Linkage*
No 4885 608 12.45 349 57.40 261 42.93 290 47.70
Yes 802 723 90.15 493 68.19 432 59.75 436 60.30
Histology
Non-seminoma/mixture 2258 584 25.86 381 65.24 309 52.91 325 55.65
Seminoma only 3353 730 21.77 448 61.37 373 51.10 390 53.42
missing 76 17 22.37 13 76.47 11 64.71 11 64.71
Race/ethnicity
Non-Hispanic white 4587 1107 24.13 715 64.59 591 53.39 621 56.10
Non-Hispanic black 59 10 16.95 6 60.00 6 60.00 6 60.00
Non-Hispanic other 253 34 13.44 19 55.88 13 38.24 13 38.24
Hispanic 788 180 22.84 102 56.67 83 46.11 86 47.78
Year of birth
<1945 640 99 15.47 69 69.70 65 65.66 62 62.63
1945-1954 1309 248 18.95 170 68.55 151 60.89 157 63.31
1955-1964 1991 486 24.41 304 62.55 257 52.88 266 54.73
1965-1974 1243 327 26.31 207 63.30 153 46.79 166 50.76
1975-1984 438 145 33.11 80 55.17 58 40.00 64 44.14
1985-1994 53 24 45.28 11 45.83 9 37.50 10 41.67
>=1995 13 2 15.38 1 50.00 0 0.00 1 50.00
Total 5678 1331 23.44 842 63.26 693 52.07 726 54.55
Phase 2 Outcomes
Invited to Phase 2 Consent Returned FHI Completed Biospecimens Received
26

Table 7. Association between bilateral, cryptorchidism, family history of TC or related conditions (linkage below), CCR-reported variables and
response to phase 2 requests: consent, family history interview, biospecimens sample.

*: Bilateral, Cryptorchidism, Linkage (family history of TC or related conditions) were self-reported from FHQ.
Adjusted model: fully mutually adjusted.
E.g.: association between bilateral and phase 2 response was adjusted for cryptorchidism, linkage, race/ethnicity, histology, year of
birth.
OR and 95% CI P>|z| N Overall P OR and 95% CI P>|z| OR and 95% CI P>|z| N Overall P OR and 95% CI P>|z| OR and 95% CI P>|z| N Overall P OR and 95% CI P>|z|
Bilateral*
No 1(ref) 1(ref) 1(ref) 1(ref) 1(ref) 1(ref)
Yes 0.75 ( 0.55- 1.01) 0.061 1331 0.0629 0.77 ( 0.55- 1.07) 0.114 0.92(0.68-1.23) 0.568 1331 0.5676 1.06 ( 0.76- 1.47) 0.739 0.87 ( 0.65- 1.17) 0.370 1331 0.3700 0.91 ( 0.66- 1.26) 0.575
Cryptorchidism*
No 1(ref) 1(ref) 1(ref) 1(ref) 1(ref) 1(ref)
Yes 0.74 ( 0.57- 0.96) 0.023 1331 0.0233 0.79 ( 0.59- 1.05) 0.105 0.89(0.67-1.12) 0.282 1331 0.2819 1.02 ( 0.77- 1.36) 0.881 0.79 ( 0.61- 1.02) 0.073 1331 0.0728 0.86 ( 0.65- 1.14) 0.29
Linkage*
No 1(ref) 1(ref) 1(ref) 1(ref) 1(ref) 1(ref)
Yes 1.59 ( 1.27- 1.99)<0.001 1331 <0.0001 1.41 ( 1.10- 1.82) 0.007 1.97(1.59-2.46) <0.001 1331 <0.0001 1.91 ( 1.49- 2.44) <0.001 1.67 ( 1.34- 2.07)<0.001 1331 <0.0001 1.53 ( 1.20- 1.95) 0.001
Race/ethnicity
Non-Hispanic white 1(ref) 1(ref) 1(ref) 1(ref) 1(ref) 1(ref)
Non-Hispanic black 0.82 ( 0.23- 2.93) 0.763 1.51 ( 0.37- 6.18) 0.568 1.31 ( 0.37- 4.67) 0.677 2.23 ( 0.54- 9.17) 0.266 1.17 ( 0.33- 4.18) 0.805 2.03 ( 0.50- 8.30) 0.325
Non-Hispanic other 0.69 ( 0.35- 1.38) 0.299 1331 0.1745 0.77 ( 0.37- 1.58) 0.470 0.54 ( 0.27- 1.09) 0.086 1331 0.1021 0.64 ( 0.31- 1.33) 0.230 0.48 ( 0.24- 0.98) 0.043 1331 0.0431 0.50 ( 0.24- 1.04) 0.063
Hispanic 0.72 ( 0.52- 0.99) 0.041 0.75 ( 0.54- 1.04) 0.088 0.75 ( 0.54- 1.02) 0.070 0.81 ( 0.59- 1.13) 0.214 0.72 ( 0.52- 0.98) 0.038 0.77 ( 0.56- 1.06) 0.112
Histology
Non-seminoma/mixture 1(ref) 1(ref) 1(ref) 1(ref) 1(ref) 1(ref)
Seminoma only 0.85 ( 0.68- 1.06) 0.149 1314 0.1482 0.74 ( 0.58- 0.94) 0.014 0.93 ( 0.75- 1.16) 0.513 1314 0.5129 0.76 ( 0.60- 0.96) 0.024 0.91 ( 0.73- 1.14) 0.421 1314 0.4207 0.79 ( 0.62- 0.99) 0.045
Year of birth
<1945 1(ref) 1(ref) 1(ref) 1(ref) 1(ref) 1(ref)
1945-1954 0.95 ( 0.57- 1.57) 0.835 0.99 ( 0.59- 1.67) 0.984 0.81 ( 0.50- 1.33) 0.408 0.87 ( 0.53- 1.44) 0.588 1.03 ( 0.64- 1.67) 0.906 1.09 ( 0.66- 1.78) 0.744
1955-1964 0.73 ( 0.46- 1.16) 0.179 0.71 ( 0.44- 1.15) 0.163 0.59 ( 0.37- 0.92) 0.021 0.60 ( 0.38- 0.95) 0.030 0.72 ( 0.46- 1.13) 0.150 0.72 ( 0.46- 1.14) 0.167
1965-1974 0.75 ( 0.46- 1.22) 0.244 1331 0.0598 0.72 ( 0.44- 1.20) 0.213 0.46 ( 0.29- 0.73) 0.001 1329 <0.0001 0.49 ( 0.30- 0.79) 0.004 0.62 ( 0.39- 0.98) 0.039 1331 0.0025 0.62 ( 0.38- 1.01) 0.053
1975-1984 0.54 ( 0.31- 0.92) 0.023 0.50 ( 0.28- 0.89) 0.017 0.35 ( 0.20- 0.59) 0.000 0.37 ( 0.21- 0.64) <0.001 0.47 ( 0.28- 0.80) 0.005 0.48 ( 0.28- 0.84) 0.009
1985-1994 0.37 ( 0.15- 0.91) 0.031 0.32 ( 0.13- 0.84) 0.02 0.31 ( 0.12- 0.79) 0.014 0.34 ( 0.13- 0.88) 0.027 0.43 ( 0.17- 1.06) 0.066 0.42 ( 0.17- 1.09) 0.074
>=1995 0.43 ( 0.03- 7.18) 0.561 0.49 ( 0.03- 9.03) 0.634 -- -- -- -- 0.60 ( 0.04- 9.83) 0.718 1.22 ( 0.07- 22.38) 0.892
Consent Returned FMI Completed Biospecimens Received
crude fully adjusted fully adjusted fully adjusted crude crude
27

rates appeared to be buried in the crude model but is apparent after controlling (p=0.014 for
responding to consent, p=0.024 for responding to FHI, p=0.045 for responding to biospecimens
request), and men with seminoma only were less likely to respond to phase 2 requests. The
following associations remained significant after adjusting: men born from 1975 through 1994
were less likely to respond to consents compared with men born before 1945 (reference group);
men born in 1995 or later were less likely to complete the FHI compared with the reference
group; and men born from 1975 through 1984 were less likely to respond to
biospecimens requests compared with the reference group.
28

Discussion
To my knowledge, the present study is the only large, population-based study of TC in
which determinants of participation were investigated. Statistical analyses identified numerous
factors to be associated with both a man’s decision to complete the screening phase of the study
and for those invite to participate further to be fully enrolled. These results have potential
implications for interpretation of estimates of risk-factor disease associations and for
generalizability of results of epidemiologic studies of TC. Moreover, they may provide valuable
insights of more general relevance to the future design of epidemiologic research addressing
conditions of adolescents and young adults.
Among 14,880 invitees to phase 1, 5687 (38.22 %) participants returned the FHQ.
Statistically significant associations were observed before and after adjustment between
responses to FHQ and each of the following variables: race/ethnicity, laterality, histology, year
of birth, diagnosis age, age at 1st mailing, and lag year. Men who were most likely to respond to
the phase 1 FHQ possessed the qualities of non-Hispanic white ethnicity, bilateral TC, seminoma
only histology (less severe), older age, older age at diagnosis, older age at time of contact by this
study, and less lag years.
In general, it appeared that older age at diagnosis and at time of contact and less lag years
were associated with more active confirmed responses to study participation, whether it be in
participation or denial. The active respondents (returned submitted questionnaire after only one
mailing) of the FHQ were more likely to be men with bilateral TC, older age, older age at
diagnosis, and older age at time of contact, and the active denials (directly communicated desire
to forgo participation) to the FHQ tended to be men with non-Hispanic white ethnicity, older
age, older age at diagnosis, older age at time of contact by this study. One would expect men
29

with less lag years (time between their receipt of study materials and return of completed FHQ)
to be more likely to respond, which proved true in this study up until 25 lag years. However, for
men whose lag years were greater than 25 years, they were actually the most likely to respond,
although chances are extremely likely that these results were confounded by other time variables,
particularly year of birth.
Among 5687 respondents who completed phase 1, 1331 (23.44%) were invited to phase
2, and among these invitees, 842 (63.26%) agreed to the consent request, 693 (52.07%)
participated in the FHI request, and 726 (54.55%) cooperated in the biospecimens request.
Statistically significant associations were found between response to phase 2 requests and
histology after adjustment. Additionally, significant positive associations were found between
response to phase 2 requests and family history of TC or related conditions, which could be
explained by concern and care about their family members’ health or their desire to understand
their own genetic predispositions to the diseases.
Our study had many qualities that enabled investigators to achieve strong results. To our
knowledge, this was the only large population-based study of TC patients for which response
rates were ever recorded in terms of demographic, medical, and family history statuses. Using
data from the California Cancer Registry, this population-based study covered all reported cases
of testicular germ cell tumor patients in California, thereby reflecting the true association of
factors affecting disease as closely as possible. Having a true population-based study allowed
investigators to have a more meaningful number of cases for conditions such as bilateral
testicular cancer, cryptorchidism, and hypospadias. The large sample size and small number of
missing data further strengthened the power of this study. The state of California is also home to
one of the largest and most ethnically and socioeconomically diverse populations in the country,
30

which offered this study the advantage of capturing a varied and comprehensive TC case
population. Lastly, multiple phases and points of contact in the study allowed investigators to
understand and cater to a participant’s preference to different modes of communication. Postal
mail, telephone calls, emails, and local in-person meetings were all used to encourage
participants to complete phase 2 requests of the study.
Despite the strengths of this study, the results must be interpreted with several
vulnerabilities in mind. One possible weakness of this study was the potential selection bias.
Since this was a prevalence study, many case listings were for individuals diagnosed decades
before attempted enrollment, CCR-reported addresses could have been out of date, causing the
men to never have received study materials, or the men, particularly the sickest, could have
already been deceased. For silent denials, those that never received the mailed study materials
should have been excluded from the total number of phase 1 participants, but there was no way
to know if this was the case. If these men truly did not receive study materials due to wrong
addresses or death, true values of response rate for phase 1 would be higher than estimates
reported here. In addition, it was not possible to fully address differing degrees of effort that may
have been exerted by investigators to enroll men in subsets of particular interest, such as those
with bilateral TC.
A limitation that could be addressed in future studies was the restricted set of variables
reported by CCR. It is reported that TC survivors suffer from impaired qualities of life and stress
from relationships, especially those in younger age groups (Smith, et al., 2013). A Swedish study
added that TC survivors tend to experience crises due to their cancer diagnosis and have great
need for psychological support (Skoogh, Steineck, Johansson, Wilderäng, & Stierner, 2013). As
stated by Bender (2012), “at least one in four testicular cancer survivors has unmet needs related
31

to financial support, body image, stress, being a cancer survivor, and fear of recurrence” (p.
2738). It seem plausible that social and behavioral factors not assessed in the study could be
related to response rate. These include: socioeconomic status, education level, marital status,
employment status, and particularly quality of life; investigation of such factors may have
provided more insight on reasons for trends in response rates.
Although research has come a long way to make TC a now-curable malignancy (Albers,
et al., 2005), more comprehensive analysis needs to be conducted encompassing an individual
patient’s psychosocial qualities in order to fully understand patterns for their crucial participation
in health research studies. This would be possible if disease registries were to collect more
qualitative data, a development that may be possible in light of emerging patterns of patient
care. For example, if electronic medical records were to systematically collect high quality
variables, a pipeline will exist for this information to be abstracted into future population-based
studies.
32

Bibliography
Albers, P., Albrecht, W., Algaba, F., Bokemeyer, C., Cohn-Cedermark, G., Horwich, A., . . .
Pizzocaro, G. (2005). Guidelines on testicular cancer. European Urology, 885-94.
Bender, J. L., Gospodarowicz, M., Wiljer, D., To, M. J., Bedard, P. L., Chung, P., . . . Warde, P.
(2012). Testicular cancer survivors' supportive care needs and use of online support: a
cross-sectional survey. Supportive Care in Cancer, 20(11):2737-46.
Brydøy, M., Oldenburg, J., Klepp, O., Bremnes, R. M., Wist, E. A., Wentzel-Larsen, T., . . .
Fosså, S. D. (2009). Observational Study of Prevalence of Long-term Raynaud-Like
Phenomena and Neurological Side Effects in Testicular Cancer Survivors. Journal of the
National Cancer Institute, 1682 - 1695.
California's population in 1950. (n.d.). Retrieved from US Census Bureau:
http://www.census.gov/dmd/www/resapport/states/california.pdf
Galea, S. T. (2007). Participation Rates in Epidemiologic Studies. Annals of Epidemiology, 643-
653.
Hill, E. M. (2013). "Let's get the best quality research we can": public awareness and acceptance
of consent to use existing data in health research: a systematic review and qualitative
study. BMC medical research methodology, 13, 1, 72-72.
Huyghe, E., Matsuda, T., & Thonneau, P. (2003). Increasing incidence of testicular cancer
worldwide: a review. Journal of Urology, 5, 5-11.
Kho, M. E. (2009). Written informed consent and selection bias in observational studies using
medical records: systematic review. BMJ (Clinical research ed.), 338, 7698, b866-b866.
33

Kuligina, E., Reiner, A., Imyanitov, E. N., & Begg, C. B. (2011). Evaluating cancer
epidemiologic risk factors using multiple primary malignancies. Epidemiology, 366 -
372.
Macleod, U., & Watt, G. C. (2008). The impact of consent on observational research: a
comparison of outcomes from consenters and non consenters to an observational study.
BMC medical research methodology, 15.
Mead, G. M., & Stenning, S. P. (2007). the international germ cell consensus classification: A
new prognostic factor-based staging classification for metastatic germ cell tumours. J
Clin Oncol, 15(2):594-603.
Menard, S. W. (1995). Applied logistic regression analysis. Thousand Oaks, CA: Sage
Publications.
Oeffinger, K. C., & Tonorezos, E. S. (2011). The cancer is over, now what? Cancer, 2250 - 2257.
SEER Stat Fact Sheets: Testis Cancer. (2014). Retrieved from Surveillance, Epidemiology, and
End Results Program: http://seer.cancer.gov/statfacts/html/testis.html
Sesterhenn, I. A., & Davis, J. J. (2004). Pathology of germ cell tumors of the testis. Cancer
Control : Journal of the Moffitt Cancer Center, 374-387.
Skoogh, J., Steineck, G., Johansson, B., Wilderäng, U., & Stierner, U. (2013). Psychological
needs when diagnosed with testicular cancer: findings from a population-based study
with long-term follow-up. BJU International, 1287-1293.
Smith, A. B., Hruby, G., Gurney, H., Turner, S., Alam, M., Cox, K., . . . Stubbs, J. (2013). The
prevalence and correlates of supportive care needs in testicular cancer survivors: a cross-
sectional study. Psycho-Oncology, 22(11):2557-6.
34

Smith, A. B., King, M., Butow, P., & Olver, I. (2013). A comparison of data quality and
practicality of online versus postal questionnaires in a sample of testicular cancer
survivors. Psycho-Oncology, 22(1):233-7.
Swartout-Corbeil, D., Frey, R. J., & Oberleitner, M. (2011). Testicular Cancer. The Gale
Encyclopedia of Medicine, 4277 - 4281.
Travis, L. B., Pukkala, E., Andersson, M., Kaijser, M., Gospodarowicz, M., Joensuu, T., . . .
Storm, H. (2005). Second cancers among 40,576 testicular cancer patients: focus on long-
term survivors. Journal of the National Cancer Institute, 1354 - 1365.
Znaor, A. L.-T. (2014). International Variations and Trends in Testicular Cancer Incidence and
Mortality. European Urology, 1095.
35

Appendix A
Partial protocol of CGCTRC study
Purpose of the study
For many years testicular cancer has been the most common malignancy of young men. Rates
have been rising steadily for the past century, affecting even more men and suggesting that there
is an environmental cause that is becoming more common. Although most men now survive
testis cancer, they are at considerable risk of developing other problems related to the disease and
its treatment: secondary cancers and other serious chronic diseases, loss of fertility, and loss of
sexual function. Established risk factors for testicular cancer are a family history of testis cancer,
suggesting genetic predisposition, and a personal history of cryptorchidism. Our long term
objective is to identify one or more environmental cause whose exposure could be reduced to
prevent testis cancer. However, such causes have been hard to find, probably because they act
early in life and are hard to measure years later after a man has been diagnosed with testis cancer.
Since testis cancer tends to cluster in families, we also believe that there are inherited genetic
causes. We are first trying to find these genetic factors, since inherited genes do not usually
change over time, allowing us to make meaningful measurements even after a man’s diagnosis.
Furthermore, by knowing genetic mechanisms of testicular cancer development, we may be able
to identify environmental exposures that exacerbate or disrupt these mechanisms. This study is
conducted in the Department of Preventive Medicine, at the Keck School of Medicine of USC in
Los Angeles, California. We are conducting a large study in which we identify men with a
history of testis cancer either from cancer registries or referrals of cases with testicular cancer by
medical personnel and support groups. Referrals also include male or female cases diagnosed
36

with germ cell tumors of testicular, ovarian, or extra-gonadal origin. We first ask these
individuals to complete a questionnaire providing information about their medical history
(including history of testis cancer and cryptorchidism) and family structure. Based on this
information, we invite some of them, along with selected members of their families to also
participate by providing us with access to medical records as well as biospecimens (tissue
collected at surgery, semen, blood, and/or saliva). The vast majority of index cases are identified
by the population based California Cancer Registry (CCR), but additional individuals, including
testicular cancer cases and male and female cases diagnosed with extra gonadal or ovarian germ
cell tumors, are referred by clinicians and survivor groups.
We then plan to conduct a genome-wide search for susceptibility genes for testis cancer and
cryptorchidism, by conducting laboratory analysis of specimens collected from study
participants, followed by statistical analyses of genetic data together with disease phenotypes and
family structure data provided by study participants. We anticipate that about 5000 people will
participate in this research. This is an observational study in which questionnaire data and
biospecimens are collected; there is no treatment or randomization aspect of research.
Major research question to be addressed in this project
Testis cancer tends to cluster in families. Fathers and brothers of testis cancer patients have
increased risk of developing the disease themselves compared to individuals with no family
history. Therefore, there is a strong genetic component to testicular cancer. For this reason, we
are trying to find the genetic factors that contribute to testis cancer predisposition. Lastly, by
uncovering the genetic mechanisms of the disease, in the long term, we may be able to shed light
on environmental exposures that may disrupt or enhance such mechanisms.
37

Procedure
To search for genetic factors that may predispose to germ cell tumors and related conditions. We
are conducting a large study in which we identify men with a history of testicular germ cell
tumors either from cancer registries or referrals by medical personnel and support groups.
Referrals will include male patients with testicular or extra-gonadal germ cell tumors as well as
female patients diagnosed with germ cell tumors.
The physician of record is notified of the details of the study and asked whether there is any
reason the patient should not be contacted If the physician does not indicate that the patient
should not be contacted within two weeks, patients is mailed a packet containing a letter
explaining the study a screening questionnaire (Family History Questionnaire, attached) that
includes the bill of rights of participants in research on the cover, and a postage-paid return
envelope. Completed questionnaires are reviewed to identify families with multiple occurrences
of testis cancer or testis cancer and cryptorchidism, and the index cases from these families are
re-contacted and invited to participate further.
Those who are interested are offered a process of informed consent by telephone and mailed
questionnaire, and those who chose to participate further complete a telephone interview
providing a fuller family history. Based on the family structure and history of testis cancer and
cryptorchidism, Dr. Cortessis identifies family members who may potentially provide genetic
information regarding etiology, and the index case is asked to provide contact information for
these family members, along with instructions as to when and how the contact should be made
(e.g. letter, phone). Family members who are interested in participating are also offered a process
of informed consent by telephone and mailed questionnaire.
38

Participants are asked to provide a sample of blood, either using a mailed kit and phlebotomy
scheduled at a laboratory or clinic near their home, or by visit of a licensed phlebotomist to their
home, as they prefer. In instances when the participant cannot provide a blood sample, saliva is
requested and collected directly into an Oragene kit returned by mail. We have sought and
obtained permission from the USC IRB to invite adult male participants to also provide samples
of ejaculated sperm, analysis of which may reveal heterogeneity of the testis cancer phenotype.
We anticipate making this request of only a small subset of men, limited to patients referred by
clinical personnel (not the CCR).
Affected individuals are also asked to complete a medical history questionnaire (attached) and to
provide HIPAA-compliant permission for the investigators to request medical information
confirming diagnose (needed for testis cancer cases not identified by the registry and
cryptorchidism cases). The HIPAA compliant release forms will be mailed to the identified
facilities requesting that medical records related to the diagnosis and treatment of the patient’s
cancer be copied and mailed back to us.
We analyze the specimens in the laboratory, and subject the resulting measures to statistical
analysis in order to try to identify genes that may predispose to testis cancer and/or undescended
testicles.

39

Appendix B
Family History Questionnaire
40

41

42

43

44

45

46

47

APPENDIX C
Medical History Questionnaire

48

49

50

51

52

53

54

Abstract (if available)

Abstract Background: Although recent decreasing response rates in epidemiologic studies have been shown to be related to factors such as race/ethnicity and age in several studies, factors affecting response rates of testicular cancer (TC) patients have barely been studied. ❧ Goal: The goal of this project was to assess factors related to a man’s decision to respond to screening (complete and return Family History Questionnaire (FHQ)) and enrollment (completing consent process, family history interview (FHI), and biospecimens donation) among TC patients in the Causes of Germ Cell Tumors and Related Conditions study (CGCTRC). ❧ Method: Case listings were retrieved from the California Cancer Registry

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

An assessment of necrosis grading in childhood osteosarcoma: the effect of initial treatment on prognostic significance

PDF

Development and validation of survey instrument designed for cervical cancer screening in Malawi, and other low resource settings

PDF

Evaluation of factors influencing Los Angeles Tiered-Dispatch System’s improvement on bystander CPR rate and inter reliability between electronic patient care report (ePCR) and 911 call review on...

PDF

Association of oophorectomy with progression of carotid atherosclerosis

PDF

Evaluating the use of friend or family controls in epidemiologic case-control studies

PDF

Expansion of emergency medical service regional routing for stroke patient care in the United States from 2000-2010

PDF

A comparison of three different sources of data in assessing the adolescent and young adults cancer survivors

PDF

Survival of children and adolescents with low-risk non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) treated with surgery only: an analysis of 234 patients from the Children’s Oncology Group stud...

PDF

Analysis of factors associated with breast cancer using machine learning techniques

PDF

Red and processed meat consumption and colorectal cancer risk: meta-analysis of case-control studies

PDF

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

PDF

Origins of the gender disparity in bladder cancer risk: a SEER analysis

PDF

Factors associated with agreement between childhood cancer survivors (CCS) and their parents on knowledge and attitudes about health-related quality of life, treatment and follow-up care

PDF

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

PDF

The effect of renal function on toxicity of E7389 (eribulin) among patients with bladder cancer

PDF

Risk factors of testicular germ cell tumor

PDF

Obesity paradox in acute heart failure decompensation

PDF

Analysis of SNP differential expression and allele-specific expression in gestational trophoblastic disease using RNA-seq data

PDF

Insulin’s effect on lactate levels in extremely low birth weight neonates. a multi-center, observational study

Asset Metadata

Creator Zhang, Yunying (author)

Core Title The study of germ cell tumors and related conditions: an analysis of self-reported data with characterization and comparison of Family History Questionnaires respondents and non-respondents

School Keck School of Medicine

Degree Master of Science

Degree Program Applied Biostatistics and Epidemiology

Publication Date 08/13/2014

Defense Date 07/01/2014

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

Tag OAI-PMH Harvest,questionnaire,response rate,testicular neoplasm

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cortessis, Victoria (committee chair), Mack, Wendy Jean (committee member), Thomas, Duncan C. (committee member)

Creator Email yunyingz@usc.edu,zhangyunyingkm@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c3-459640

Unique identifier UC11287587

Identifier etd-ZhangYunyi-2817.pdf (filename),usctheses-c3-459640 (legacy record id)

Legacy Identifier etd-ZhangYunyi-2817.pdf

Dmrecord 459640

Document Type Thesis

Format application/pdf (imt)

Rights Zhang, Yunying

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