University of Southern California Dissertations and Theses

Epidemiological studies of Epstein-Barr virus, lymphoma, and immune conditions

(USC Thesis Other)

Epidemiological studies of Epstein-Barr virus, lymphoma, and immune conditions

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

EPIDEMIOLOGICAL STUDIES OF EPSTEIN-BARR VIRUS,
LYMPHOMA, AND IMMUNE CONDITIONS

by

Niquelle Brown Wadé

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

May 2020

Copyright 2020 Niquelle Brown Wadé
ii

i have a life to garden
a multiverse to wake from sleep.

—giants

Nayyirah Waheed
iii

DEDICATION
To Naiyira and your radiant curiosity.
To Jareau and your lighthouse love.
To Raianna and your ever-giving empathy.
To Sheereen and your intuitive consistency.
To Daddy and your larger-than-life stories.
To Mommy for lighting the path.

iv

ACKNOWLEDGEMENTS
To my committee chair, Dr. Wendy Cozen, thank you for the countless hours of
mentoring and guidance. To my qualifying exam and dissertation committee members (Dr.
Thomas Mack, Dr. David Conti, Dr. Ruth Jarrett, Dr. Victoria Cortessis, Dr. Joshua
Millstein, and Dr. Clive Taylor), thank you for your thoughtful insight and encouragement.
To Dr. Christianne Lane and Dr. Yu Shen Lin, thank you for making space for my doctoral
work and advocating on my behalf. To the league of friends, professors, mentors, family
members, and colleagues whose heartening support I received during my doctoral journey,
thank you for making this dissertation possible.

v

TABLE OF CONTENTS
Dedication iii
Acknowledgements iv
List of Tables vii
List of Figures x
Abbreviations xiii
Abstract xv
Chapter 1. Introduction 1
Epstein-Barr Virus and Infectious Mononucleosis 3
Epstein-Barr Virus and Lymphoid Neoplasms in Nonimmunocompromised
Persons 10
Epstein-Barr Virus in Immunocompromised Persons 13
Epstein-Barr Virus and Immune-Associated Diseases 14
Specific Aims 15
Tables 17
Chapter 2. History of infectious mononucleosis and immune conditions: a
population-based cross-sectional study among twins 20
Abstract 20
Materials and Methods 23
Results 26
Twin pair characteristics (zygosity, sex, race/ethnicity, mother’s age at twins’
birth, number of siblings, birth order, response age, and IM status) are shown in
Tables 26
Discussion 29
Tables 34
Chapter 3. Infectious mononucleosis, immune genotypes, and non-Hodgkin
lymphoma (NHL): a pooled case-control study from the InterLymph
Consortium 37
Publication authors 37
Author affiliations 38
Abstract 40
Funding 41

vi

Disclaimers 42
Materials and Methods 45
Results 51
Discussion 53
Tables 59
Chapter 4. A genome-wide association study of Epstein-Barr virus viral copy
number and antibodies to Epstein-Barr virus antigens among Hodgkin
lymphoma survivors 73
Abstract 73
Materials and Methods 76
Results 82
Discussion 85
Tables 88
Figures 99
Supplemental Materials 104
Chapter 5. Discussion 123
Summary of findings 123
Future directions 127
References 129

vii
LIST OF TABLES
Table 1.1. Known or purported function of EBV Latent Antigens.
Adapted from Macsween and Crawford (2003) and
Sullivan (2019). 17
Table 1.2. EBV-Associated Lymphoid Malignancies in
Nonimmunocompromised Persons. Adapted from
Fields Virology 2013. 18
Table 1.3. EBV-Associated Lymphoid Malignancies in
Immunocompromised Persons. Adapted from
Longnecker, Kieff, Cohen (2013). 19
Table 2.1. Characteristics of twin pairs by IM status 34
Table 2.2. Risk of IM according to history of atopic or
autoimmune medical conditions among 465 IM-
discordant like-sex twin pairs 35
Table 2.3. Infectious mononucleosis association with atopic
conditions and autoimmune disease among 3699 like-
sex white/European American twin pairs 36
Table 3.1: Source of Participants from InterLymph Case-Control
Studies: EpiLymph 59
Table 3.2. Source of Participants from InterLymph Case-Control
Studies: BC, Mayo Clinic, NCI-SEER, Scale, UCSF,
Yale 60
Table 3.3. Demographic Characteristics of 7926 NHL Patients 61
Table 3.4. Demographic Characteristics of 10018 Controls 62
Table 3.5. Subtypes Among NHL Patients 63
Table 3.6. Associations Between NHL and Candidate Risk
Variants [IL1A (rs1800587), IL1B (rs16944,
rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6
(rs1800795, rs1800797), IL10 (rs1800896,
rs1800890), TNFA (rs1800629), HLA I (rs6457327),
and HLA II (rs10484561)] 64
Table 3.7. Association Between NHL and Infectious
Mononucleosis by NHL Subtype 65
Table 3.8. Interaction Between Infectious Mononucleosis History
and Candidate Risk Variants [IL1A (rs1800587), IL1B
(rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6
(rs1800795, rs1800797), IL10 (rs1800890), TNFA
(rs1800629), HLA I (rs6457327), and HLA II

viii
(rs10484561)] on NHL, CLL/SLL, and DLBCL risk:
Empirical-Bayes Estimates of Interaction Effects 66
Table 3.9. Interaction Between Infectious Mononucleosis History
and Candidate Risk Variants [IL1A (rs1800587), IL1B
(rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6
(rs1800795, rs1800797), IL10 (rs1800890), TNFA
(rs1800629), HLA I (rs6457327), and HLA II
(rs10484561)] on FL, MCL, and TCL risk: Empirical-
Bayes Estimates of Interaction Effects 67
Table 3.10. Association Between Infectious Mononucleosis
History and T-Cell Lymphoma Among Genotyped
Participants Stratified by IL1B (rs16944, rs1143627)
and IL6 (rs1800795, rs1800797) Genotypes 68
Table 3.11. Association Between Age at NHL Diagnosis and
Infectious Mononucleosis 69
Table 3.12. Associations Between Infectious Mononucleosis
History and Demographic Factors Among 10018
Controls 70
Table 3.13. Association Between Infectious Mononucleosis
History and Family Structure Among Controls with
Available Year of Birth Data 71
Table 3.14. Associations Between Infectious Mononucleosis and
Candidate Risk Variants [IL1A (rs1800587), IL1B
(rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6
(rs1800797), IL10 (rs1800890), TNFA (rs1800629),
HLA I (rs6457327), and HLA II (rs10484561)]
Among Controls with Available Genetic Data 72
Table 4.1. Characteristics of Participants from USC and UPRD
Discovery Sets 88
Table 4.2. Pre- and Post-Imputation Availability of Genotyping
Data for Study Participants 89
Table 4.3. Associations Between Chromosome 3 Variants and
EBV Viral Load Among 166 Participants with
Hodgkin Lymphoma in USC Discovery Set 90
Table 4.4. Associations Between Chromosome 3 Variants and
EBV Viral Load Among 106 Participants with
Hodgkin Lymphoma in UPRD Discovery Set 91
Table 4.5. Associations Between Chromosome 3 Variants and
EBV Viral Load Among 272 Participants with
Hodgkin Lymphoma in Meta-Analysis 92

ix
Table 4.6. ANNOVAR Functional Annotation of rs2246901 93
Table 4.7. Association Between the A Allele of rs2246901 and
EBV Viral Load Among 272 Hodgkin Lymphoma
Survivors Stratified by Discovery Set, Histology,
Tumor EBV Status, and Infectious Mononucleosis
History 94
Table 4.8. Association Between the A Allele of rs2246901 and
EBV Viral Load Among 272 Hodgkin Lymphoma
Survivors Stratified by Sex, Age at Diagnosis, Age at
Blood Draw, and Time between Diagnosis and Blood
Draw 95
Table 4.9. Variants Most Strongly Associated with MFI Levels of
Antibodies to EBV Antigens Among USC Participants
Exceeding MFI Threshold for Antibody Positivity 95
Table 4.10. Association Between the C Allele of rs115805790 and
Levels of Antibodies to EBNA-1 Antigen Among 124
USC Participants with EBNA-1 Levels Surpassing 411
MFI by Sex, Histology, Tumor EBV Status, and Self-
Reported Infectious Mononucleosis History 97
Table 4.11. Association Between the A Allele of rs140444865 and
Levels of Antibodies to EBNA-1 Antigen Among 124
USC Participants with EBNA-1 Levels Surpassing 411
MFI by Sex, Histology, Tumor EBV Status, and Self-
Reported Infectious Mononucleosis History 98
Supplemental Table 4.1. Characteristics of 143 Hodgkin Lymphoma Survivors
from USC with Sufficient Blood Sample for Analysis
of Antibodies to EBV Antigens 104

x
LIST OF FIGURES
Figure 1.1. Lytic and Latent Phases of Epstein-Barr Virus
Infection. Adapted from Odumade, Hogquist, and
Balfour (2011). 3
Figure 1.2. CD8+ and CD4+ T-cell Responses to Epstein-Barr
Virus Lytic [Immediate-Early (IE), Early (E), Late
(L)] and Latent Cycle Proteins. Adapted from Hislop
et al. (2007). 4
Figure 1.3: EBV-specific antibodies and viral load following EBV
infection. Adapted from Odumade et al. (2011). 7
Figure 1.4. EBV Seroconversion by Age for Selected Countries.
Adapted from Hjalgrim et al. (2007). 9
Figure 1.5. Four Disease Model of Hodgkin Lymphoma. Adapted
from Jarrett (2002) and Macsween and Crawford
(2003). 11
Figure 4.1. Genome-Wide Association Manhattan Plot for EBV
Viral Load: Results from Meta-Analysis 99
Figure 4.2. LocusZoom Plot for Associations Between MUC4
Variants in Chromosome 3 and EBV Viral Load:
Results from Meta-Analysis 100
Figure 4.3. Genome-Wide Association Manhattan Plot for Median
Fluorescence Intensity (MFI) of Antibodies to Epstein-
Barr Virus Nuclear Antigen 1 (EBNA-1) Among
Participants Exceeding MFI Threshold for EBNA-1
Positivity: Results from University of Southern
California Discovery Set 101
Figure 4.4. LocusZoom Plot for Association Between
rs115805790 (INHA;STK11IP) in Chromosome 2 and
Median Fluorescence Intensity (MFI) of Antibodies to
Epstein-Barr Virus Nuclear Antigen 1 (EBNA-1)
Among Participants Exceeding MFI Threshold for
EBNA-1 Positivity: Results from University of
Southern California Discovery Set 102
Figure 4.5. LocusZoom Plot for Association Between
rs140444865 (GRM7) in Chromosome 3 and Median
Fluorescence Intensity (MFI) of Antibodies to Epstein-
Barr Virus Nuclear Antigen 1 (EBNA-1) Among
Participants Exceeding MFI Threshold for EBNA-1

xi
Positivity: Results from University of Southern
California Discovery Set 103
Supplemental Figure 4.1. Distribution of EBV Viral Copy number by Participant
Characteristics 105
Supplemental Figure 4.2. Quantile-Quantile Plots (-log10 Scale) for EBV Viral
Load: Results from University of Southern California
and Université de Paris René Descartes Discovery Sets 106
Supplemental Figure 4.3. Quantile-Quantile Plot (-log10 Scale) for EBV Viral
Load: Results from Meta-Analysis 107
Supplemental Figure 4.4. Distribution of Median Fluorescence Intensity (MFI)
of Antibodies to Early Antigen-Diffuse (EA-D) by
Participant Characteristics 108
Supplemental Figure 4.5. Distribution of Median Fluorescence Intensity (MFI)
of Antibodies Epstein-Barr Virus Nuclear Antigen 1
(EBNA-1) by Participant Characteristics 109
Supplemental Figure 4.6. Distribution of Median Fluorescence Intensity (MFI)
of Antibodies Viral Capsid Antigen p18 (VCA p18) by
Participant Characteristics 110
Supplemental Figure 4.7. Distribution of Median Fluorescence Intensity (MFI)
of Antibodies Z-Epstein-Barr Virus Replication
Activator (ZEBRA) by Participant Characteristics 111
Supplemental Figure 4.8. Genome-Wide Association Manhattan Plot for
Presence of Antibodies to Early Antigen-Diffuse (EA-
D): Results from University of Southern California
Discovery Set 112
Supplemental Figure 4.9. Genome-Wide Association Manhattan Plot for the
Presence of Antibodies to Epstein-Barr Virus Nuclear
Antigen 1 (EBNA-1): Results from University of
Southern California Discovery Set 113
Supplemental Figure 4.10. Genome-Wide Association Manhattan Plot for the
Presence of Antibodies to Viral Capsid Antigen p18
(VCA p18): Results from University of Southern
California Discovery Set 114
Supplemental Figure 4.11. Genome-Wide Association Manhattan Plot for the
Presence of Antibodies to Z-Epstein-Barr Virus
Replication Activator (ZEBRA): Results from
University of Southern California Discovery Set 115
Supplemental Figure 4.12. Genome-Wide Association Manhattan Plot for Serum
EBV Positivity: Results from University of Southern
California Discovery Set 116

xii
Supplemental Figure 4.13. Quantile-Quantile Plots(-log10 Scale) for Presence of
Antibodies to EBV Antigens: Results from University
of Southern California Discovery Set 117
Supplemental Figure 4.14. Quantile-Quantile Plot (-log10 Scale) for Serum EBV
Positivity: Results from University of Southern
California Discovery Set 118
Supplemental Figure 4.15. Genome-Wide Association Manhattan Plot for Median
Fluorescence Intensity (MFI) of Antibodies to Early
Antigen-Diffuse (EA-D) Among Participants
Exceeding MFI Threshold for EA-D Antibody
Positivity: Results from University of Southern
California Discovery Set 119
Supplemental Figure 4.16. Genome-Wide Association Manhattan Plot for Median
Fluorescence Intensity (MFI) of Antibodies to Viral
Capsid Antigen p18 (VCA p18) Among Participants
Exceeding MFI Threshold for VCA p18 Positivity:
Results from University of Southern California
Discovery Set 120
Supplemental Figure 4.17. Genome-Wide Association Manhattan Plot for Median
Fluorescence Intensity (MFI) of Antibodies to Z-
Epstein-Barr Virus Replication Activator (ZEBRA)
Among Participants Exceeding MFI Threshold for
VCA ZEBRA Positivity: Results from University of
Southern California Discovery Set 121
Supplemental Figure 4.18. Quantile-Quantile Plots (-log10 Scale) for Median
Fluorescence Intensity of Antibodies to EBV
Antigens: Results from University of Southern
California Discovery Set 122

xiii
ABBREVIATIONS
BC British Columbia
BL Burkitt lymphoma
CI confidence interval
CLL chronic lymphocytic leukemia
CTP California Twin Program
DLBCL diffuse large B-cell lymphoma
DZ dizygotic
EA European American
EA-D early antigen-diffuse
EBNA Epstein-Barr virus nuclear antigen
EBV Epstein-Barr virus
FL follicular lymphoma
GC germinal center
GWAS genome-wide association study
HHV human herpesvirus
HL Hodgkin lymphoma
HLA human leukocyte antigen
IgE Immunoglobulin E
IL interleukin
IM infectious mononucleosis
LD lymphocyte depleted
LMP latent membrane protein
LP leader protein
MAF minor allele frequency
MC mixed cellularity
MCL mantle cell lymphoma
MF/SS mycosis fungoides/Sézary syndrome
MFI median fluorescence intensity
MZ monozygotic

xiv
NCI-SEER National Cancer Institute Surveillance, Epidemiology, and End
Results
NHL non-Hodgkin lymphoma
NOS not otherwise specified
NS nodular sclerosis
OR odds ratio
PBon Bonferroni-adjusted P value
PLL prolymphocytic lymphoma
SCALE Scandinavian lymphoma etiology
SD standard deviation
SES socioeconomic status
SLL small lymphocytic lymphoma
TCL T-cell lymphoma
TH2 T helper type 2
TNF tumor necrosis factor
UCSF University of California San Francisco
UPRD Université de Paris René Descartes
USC University of Southern California
VCA viral capsid antigen
ZEBRA Z-Epstein-Barr virus replication activator

xv
ABSTRACT
Epstein-Barr virus (EBV) is a ubiquitous human herpes virus which is estimated
to infect over 90% of adults worldwide. Infectious mononucleosis (IM), a clinical
syndrome typified by fever, pharyngitis, and lymphadenopathy, is a response to primary
EBV infection. IM is most common among individuals whose primary EBV infection
occurs during adolescence or young adulthood. In addition to serious sequelae related to
IM, EBV is associated with several malignancies. EBV DNA is present in 95–100% of
Burkitt lymphomas in endemic regions, 100% of anaplastic nasopharyngeal carcinomas,
60–90% of mixed cellularity and lymphocyte-depleted Hodgkin lymphoma subtypes,
100% of lymphoepitheliomas, and 25–100% of various non-Hodgkin lymphoma
subtypes. By conservative estimates, at least 113,000 incident cases of cancer each year,
roughly 1% of worldwide cancer incidence, are attributable to EBV infection and could
be prevented if EBV was eliminated or successfully controlled by the immune system.
Despite the established associations with immunopathology and malignancy, questions
remain about the role of environmental, genetic, and immunological factors in
determining the severity of primary EBV infection, the implications of symptomatic
infection for future disease risk, and the role of genetic variation in symptomatic primary
and reactivated (replicating) EBV infection. This dissertation includes three projects
aimed at better framing the relationship between EBV, IM, and other immune-related
conditions including autoimmune disease, atopy, Hodgkin lymphoma, and non-Hodgkin
lymphoma.

1
Chapter 1. INTRODUCTION
Infectious mononucleosis (IM), a clinical syndrome typified by fever, pharyngitis,
and lymphadenopathy, is a response to primary Epstein-Barr virus (EBV) infection. IM is
most common among individuals whose primary EBV infection occurs during
adolescence or young adulthood (1). EBV is a ubiquitous human herpes virus responsible
for 500 cases of IM per 100,000 people in the United States each year (2,3). In addition to
the strain on productivity caused by IM-related fatigue, which can last six months or
longer in some patients, 1% of those diagnosed with IM will develop serious
neurological, hepatological, or hematological sequelae (4). Splenomegaly occurs in 50–
60% of IM patients and can result in splenic rupture, a rare but potentially fatal
consequence of IM (5).
Along with these serious sequelae related to primary infection, EBV is associated
with several malignancies. EBV DNA is present in 95–100% of Burkitt lymphomas in
endemic regions, 100% of anaplastic nasopharyngeal carcinomas, 60–90% of mixed
cellularity and lymphocyte-depleted Hodgkin lymphoma subtypes, 100% of
lymphoepitheliomas, and 25–100% of various non-Hodgkin lymphoma (6). By
conservative estimates, at least 113,000 incident cases of cancer each year, roughly 1% of
worldwide cancer incidence, are attributable to EBV infection and could be prevented if
EBV was eliminated or successfully controlled by the immune system (7).
A weaker body of evidence exists for associations between EBV and autoimmune
diseases. Associations between rheumatoid arthritis and elevated EBV viral loads have
been reported (8) as have associations between systemic lupus erythematosus and

2
elevated antibody titers to EBV (9). Lastly, a combined history of IM, elevated antibody
titers to EBV, and elevated EBV specific T cells are all associated with risk of multiple
sclerosis (10–12).
Despite the established associations with immunopathology and malignancy,
questions remain about the role of environmental, genetic, and immunological factors in
determining the severity of primary EBV infection, the implications of symptomatic
infection for future disease risk, and the role of genetic variation in symptomatic primary
and reactivated (replicating) EBV infection. EBV is a pervasive virus which is well-
adapted to human hosts. These features alongside known associations with various
malignancies and autoimmune diseases make it an epidemiologically interesting subject
worthy of further inquiry.
This dissertation includes three projects aimed at better framing the relationship
between EBV, IM, and other immune-related conditions including autoimmune disease,
atopy, Hodgkin lymphoma, and non-Hodgkin lymphoma. In the first project, we used a
cross-sectional approach to explore associations between self-reported infectious
mononucleosis, atopy, and autoimmune conditions among a population-based sample of
twins. In the second project, a pooled case-control study was used to [1] examine
potential interaction between self-reported IM history and a candidate panel of 12
immune-related genetic variants on risk of non-Hodgkin lymphoma and [2] explore
potential risk factors for IM among controls. In the third project, another cross-sectional
study, we assessed the relationship between genetic variation, EBV viral load, and
antibodies to EBV antigens among Hodgkin lymphoma patients.

3
EPSTEIN-BARR VIRUS AND INFECTIOUS MONONUCLEOSIS
Virology of Epstein-Barr virus
EBV is the common name for human herpesvirus 4 (HHV-4), a
lymphocryptovirus that has evolved with our species to achieve a latent infection in
human hosts after an often asymptomatic primary infection (6). Membership to the
herpes family of viruses is determined by virion morphology. Each herpesvirus consists
of a spherical virion with a double-stranded DNA genome, capsid, tegument, and
envelope (13). Like the other 200+ herpesviruses identified to date, the nine human
herpesviruses maintain lifelong latent infection in hosts with periodic viral shedding (14).

Figure 1.1. Lytic and Latent Phases of Epstein-Barr Virus Infection. Adapted from
Odumade, Hogquist, and Balfour (2011).
EBV achieves episomal latency by preferentially targeting lymphocytes. Despite
this lymphotropic behavior, the initial site of infection is in squamous epithelial cells in
the oropharynx (15). This lytic phase of EBV infection is marked by high levels of viral

4
shedding in the throat which infects nearby resting B lymphocytes in Waldeyers ring—
mucosa-associated lymphoid tissue including the palatine, pharyngeal, and lingual
tonsils. The cycle between lysis and latency is illustrated in Figure 1.1 adapted
from Odumade, Hogquist, and Balfour (16).

Figure 1.2. CD8+ and CD4+ T-cell Responses to Epstein-Barr Virus Lytic [Immediate-
Early (IE), Early (E), Late (L)] and Latent Cycle Proteins. Adapted from Hislop et al.
(2007).
EBV proteins can be classified into four groups according to the timing of their
presentation: immediate-early (IE) lytic, early (E) lytic, late (L) lytic, and latent cycle.
CD8+ and CD4+ T-cell responses to these proteins are shown in Figure 1.2, adapted from
Hislop et al. (17). In the figure, proteins which had not been tested are denoted with n.t.,
and dotted lines represent CD4+ T-cell responses of unknown immunodominance. In
general, IE, E, and latent proteins have the strongest impact on CD4+ and CD8+ T-cell
responses. The roles of specific latent cycle proteins are described in Table 1.1, adapted
from Macsween and Crawford (18) and Sullivan (19).

5
T-cell response to primary EBV infection
Primary EBV infection can either present clinically as IM or result in subclinical
seroconversion. During IM, EBV viral load is elevated in the blood, and CD8+ T-cell
count is dramatically elevated due to an amplified EBV-specific CD8+ T-cell response
(20). The majority of CD8+ T-cells correspond to specific proteins produced during the
EBV lytic cycle (predominantly IE proteins and, to a lesser degree, E proteins). L
proteins are seldom targeted by this T-cell response (21). CD8+ T-cells for specific latent
cycle EBV proteins (LMP2 and EBNA-3A, 3B, and 3C) are present, but they appear less
frequently (22,23). The inverse is true for CD4+ T-cell response during IM; latent antigen
responses outnumber lytic antigen responses. Although the magnitude of the CD4+ T-cell
response to EBV in IM is smaller than that of the CD8+ T-cell response, the CD4+
response has a broader range of targets (24,25). Upon resolution of IM symptoms, EBV
viral load along with EBV-specific CD8+ and CD4+ T-cells decrease to stable values
indicative of persistent, lifelong infection (24,26,27).
Although IM is an uncommon response to primary EBV infection, the majority of
studies of EBV seroconversion have been conducted among IM patients (25). Subclinical
primary infection is most common in childhood and is often undetected or asymptomatic.
Studies of asymptomatic seroconversion are more difficult to undertake because they
require close, longitudinal monitoring of seronegative individuals. Prospective studies of
asymptomatic primary infection in both children and young adults have shown high
levels of EBV DNA in the blood and a high percentage of EBV-specific CD8+ T-cells.
No change in the overall size of the T-cell compartment has been observed (28–31).
These findings suggest IM is, at least in part, caused by the host’s T-cell response.

6
Genetic susceptibility to infectious mononucleosis
Studies indicating familial IM clustering and higher concordance for IM risk
among monozygotic compared to dizygotic twin pairs suggest a role for genetic
susceptibility in IM etiology (32,33). Associations have been reported between IM
(development and severity) and polymorphisms in the human leukocyte antigen (HLA)
class I and II regions, which may affect immune response by altering the efficiency of
presentation of viral peptides to T-cells (34–36). Associations between polymorphisms in
the IL10 gene, which codes for an anti-inflammatory cytokine by the same name, and
severity/age of EBV seroconversion have been reported as well (37,38).
Infectious mononucleosis signs, symptoms, and severity
The most common signs and symptoms of IM are lymphadenopathy (100%
frequency), malaise and fatigue (90–100%), fever (80–95%), pharyngitis (85%), and
sweats (80–95%) (4). During typical adolescent presentation of IM, the onset of
lymphadenopathy, fever, and pharyngitis are often presaged by malaise, myalgia, and/or
fatigue (39). While most symptoms resolve within one month of diagnosis, many
participants continue to experience fatigue. Evidence suggests this prolonged fatigue
(>180 days) may be more common in females than males (40). Few pediatric cases of
primary EBV infection are diagnosed as IM. In addition to the higher frequency of
asymptomatic primary EBV infection, this lack of pediatric IM diagnoses could be due to
smaller doses of viral inoculum, milder symptomology in childhood, and insensitivity of
heterophile antibody tests among young children (39,41,42).

7
Diagnosis of infectious mononucleosis and primary Epstein-Barr virus infection
Because IM signs and symptoms vary and can be caused by other conditions,
these indicators are neither sensitive nor specific means of diagnosing primary EBV
infection (43). Serologic tests provide a more accurate method for confirming a patient is
experiencing IM symptoms as a result of primary EBV infection.

Figure 1.3: EBV-specific antibodies and viral load following EBV infection. Adapted
from Odumade et al. (2011).
Heterophile antibody tests, pioneered by Paul and Bunnell in 1932, offer a low-
cost sensitive method for confirming suspected cases of IM caused by EBV in adults, but
are not sensitive during the first week of adult EBV infection or in children experiencing
primary EBV infection (44). EBV-specific antibody tests are often used for follow-up in

8
patients with suspected IM who are heterophile negative. As shown in Figure 1.3 adapted
from Odumade et al. (16), immunoglobulin G (IgG) and M (IgM) antibodies to EBV viral
capsid antigen (VCA) are typically present at IM symptom onset. While EBV IgG VCA
antibodies persist at steady levels for life, EBV IgM VCA antibody levels peak within a
month of symptom onset and dwindle to undetectable levels within two to six months
following infection. EBNA-1 (Epstein-Barr virus nuclear antigen-1) IgG antibodies
develop three to six months after infection onset. An almost unequivocal diagnosis of
acute EBV infection can be made in the presence of IgM VCA and the absence of IgG
EBNA antibodies (5).
Infectious mononucleosis epidemiology
The probability of clinical IM diagnosis is elevated among those whose primary
EBV infection occurs after the age of 10 years. Although measurements of IM incidence
vary significantly based on the degree of clinical follow-up (4,26), it is generally
accepted that roughly 25% of primary EBV infections occurring in adolescence and
young adulthood will result in IM (6). Because affluence is associated with decreased
exposure to common childhood infections, protection from early-life EBV exposure is
common among children in households and geographies of higher socioeconomic status
(45,46). In Figure 1.4, adapted from Hjalgrim et al. (2007), we observe early infection, a
decline in early-life antibody positivity corresponding to diminishing availability of
maternal antibodies, a lack of new infections in school-aged children, and a subsequent
increase in infections during adolescence (46). In non-industrial countries, EBV
seropositivity is typically universal by age four (4). Studies in the US have shown 50%
seropositivity by age 2 in areas of low socioeconomic status and 39% seropositivity

9
among all children ages 5 to 14 (29,47). Worldwide, 90–95% of all adults are EBV-
positive (47–50).

Figure 1.4. EBV Seroconversion by Age for Selected Countries. Adapted from Hjalgrim
et al. (2007).
EBV is spread through salivary exchange of epithelial cells, and those who evade
primary EBV infection during childhood are likely to be infected during adolescence
when deep kissing becomes common (51,52). Historical studies have shown IM
incidence peaks at 10 to 19 years of age in industrialized countries (6–8 cases per 1000
person-years). Studies performed in Japan and the UK also indicate later exposure to
EBV results in heightened frequency and severity of clinical IM diagnoses (53,54).

10
EPSTEIN-BARR VIRUS AND LYMPHOID NEOPLASMS IN
NONIMMUNOCOMPROMISED PERSONS
Characteristics of EBV-associated lymphoid malignancies among
nonimmunocompromised persons are described in Table 1.2, which is adapted from
Fields Virology (6). The frequency of EBV-positive tumors varies by disease. EBV is
found in tumor cells of all cases of peripheral T-cell lymphoma associated with chronic
active or acute EBV infection, diffuse large B-cell lymphoma (DLBCL) associated with
chronic inflammation, and nasal type extranodal NK/T lymphoma. It is also found in the
B-cells of patients with angioimmunoblastic T-cell lymphoma. EBV frequency varies by
subtype for Hodgkin lymphoma and by geographic region for Burkitt lymphoma. EBV
gene expression, latency phase, and latency period after EBV infection also vary by
disease.
Hodgkin lymphoma and Epstein-Barr virus
Classic Hodgkin lymphoma (HL) is a cancer characterized by clonal tumor cells
including large, multinucleated cells and mononuclear cells —Hodgkin Reed-Sternberg
cells —in an inflammatory background (55). A substantial body of evidence suggests
EBV plays a role in the pathogenesis of a subset of HL cases (56). Compared to patients
with other lymphomas, HL patients have elevated EBV antibody titers (57) several years
before HL diagnosis (58). EBV proteins and viral DNA are present in Reed-Sternberg
cells of 40% of HL cases in the United States and Europe (59,60). In a registry-based
study of HL in Denmark, risk of EBV-positive HL was elevated in participants who had
been diagnosed with IM while risk of EBV-negative HL was not elevated after IM

11
diagnosis (61). Frequency of EBV-positivity varies significantly by HL histologic
subtype (nodular sclerosis: 20–40%; mixed cellularity: 60–75%; lymphocyte-depleted:
80–90%, lymphocyte rich: <10%) and patient characteristics (Hispanic ethnicity, sex,
socioeconomic status, immunosuppression, and age at diagnosis) (6,62). Genome-wide
association studies (GWAS) have revealed overlapping, but distinct, genetic risk profiles
for EBV-positive and EBV-negative HL (63,64). Among EBV-positive HL, EBV DNA is
detectable in plasma before HL therapy begins. Plasma EBV viral load declines in
response to therapy and increases prior to HL recurrence (65–67).

Figure 1.5. Four Disease Model of Hodgkin Lymphoma. Adapted from Jarrett (2002) and
Macsween and Crawford (2003).
A four-disease model of HL stratifies cases into four groups according to age of
incidence peak and tumor EBV status. Three of these age-specific disease strata are EBV-
associated, and the fourth (largest peak) is not. The EBV-associated strata are

12
characterized by incidence peaks below ten years of age (low- and middle-income
countries), between 15 and 34 years of age (delayed exposure to EBV), and over 55
years. Non-EBV-associated HL accounts for the majority of cases and the young adult
incidence peak observed in higher-income countries (56). This four disease HL model is
illustrated in Figure 1.5, adapted from Jarrett (56) and Macsween and Crawford (18).
Solid lines represent EBV-associated HL strata, and the dashed line represents the non-
EBV-associated stratum.
Non-Hodgkin lymphoma and Epstein-Barr virus
Non-Hodgkin lymphoma (NHL) comprises a diverse group of lymphoid
malignancies with distinct histopathologies and risk patterns originating from B- (~80%)
and T-lymphocytes (~20%) (68). EBV was first discovered in the tumor cells of patients
with Burkitt lymphoma (BL), an aggressive B-cell neoplasm (2). As with other aspects of
BL epidemiology and pathobiology, the relationship between BL and EBV varies by
geographic region. In regions where malaria is endemic, nearly all cases of BL are EBV-
positive, and it is hypothesized that malaria and EBV infections cooperate to cause BL.
Malaria infection leads to higher EBV viral load, an increased number of B cells, and
immune suppression which reduces surveillance of EBV (6,69,70). T-cell lymphomas
have been observed in persons with chronic EBV infection, and clonal EBV DNA
indicates T-cell lymphomas can be clonal expansions of a single cell infected with EBV
(71,72). A minority of DLBCLs are EBV-positive, and EBV-positivity is associated with
poorer response to treatment, older age, and extranodal involvement among DLBCL
patients (73).

13
EPSTEIN-BARR VIRUS IN IMMUNOCOMPROMISED PERSONS
Characteristics of EBV-associated malignancies among immunocompromised
persons are described in Table 1.3, which is adapted from Fields Virology (6). Those
with congenital, iatrogenic, or acquired immunodeficiency are at increased risk for EBV-
associated malignancy due to impaired T-cell function and an inability to control EBV-
infected B-cells. Among human immunodeficiency virus (HIV) patients, Burkitt and
DLBCL are the most commonly occurring lymphoma subtypes. Together, they account
for 75% of HIV lymphomas (6). The proportion of HIV-associated BL cases that are
EBV-positive (30–70%) is similar to the rate of EBV-positivity in sporadic BL cases
among non-immunocompromised individuals (15–85%) rather than that of endemic BL
cases among non-immunocompromised individuals (95–100%).
The CD4+ and CD8+ T-cell response of HIV patients on highly active
antiretroviral therapy (HAART) is similar to that of healthy individuals (74). The
introduction of HAART was associated with a decline in the overall incidence of EBV-
positive lymphomas but not lymphomas that occur early in the course of HIV (Burkitt
and Hodgkin lymphoma) (75). A notable exception to EBV-associated lymphoma risk
occurs among persons with X-linked agammaglobulinemia, a rare genetic immune
disorder in which patients lack mature B cells. These patients are not infected with EBV,
nor do they have cellular immunity to the virus (76).

14
EPSTEIN-BARR VIRUS AND IMMUNE-ASSOCIATED DISEASES
Autoimmunity and Epstein-Barr virus
Autoimmunity occurs when the immune response of a host is directed toward its
own healthy cells and tissues (77). Autoimmune disease results from systemic or organ-
specific autoimmunity and affects roughly 5% of the population in the United States and
Europe (78). Interaction between genetic susceptibility and environmental triggers
induces autoimmune response and may promote the progression of autoimmune disease.
EBV has been implicated as a possible environmental trigger for a subset of autoimmune
conditions (79).
A strong association between IM and multiple sclerosis, an inflammatory
demyelinating neurological disease, was reported in a meta-analysis and confirmed in a
registry-based longitudinal study (10,80). While EBV has been implicated as a possible
etiologic agent for systemic lupus erythematosus (9) and rheumatoid arthritis (81),
associations between these autoimmune conditions and IM have not been studied
extensively.
Atopy and Epstein-Barr virus
Atopy is a genetic predisposition to produce immunoglobulin E (IgE) antibodies
in reaction to allergen exposure due to a T helper type 2 (TH2) response from CD4+ T
helper cells (82). These TH2 cells produce interleukins 4, 5, 10, and 13 (IL-4, IL-5, IL-10,
and IL-13). IL-4 and IL-13 secreted by TH2 cells interact with B-cells to induce
production of allergen-specific IgE (83–85). A portion of individuals with allergen-
specific IgE responses also exhibit symptoms when exposed to the allergen. This

15
coupling of IgE response and symptoms manifests as allergic conditions, the most
common of which are allergic rhinitis, allergic asthma, atopic dermatitis, food allergy,
insect allergy, and drug allergy (86). Although serum IgE levels are typically higher
among individuals with allergic conditions, there is significant overlap in the range of IgE
levels among those with and without known allergic conditions (87).
Several studies have reported associations between age at EBV infection and
atopic conditions (allergies, chronic eczema, and asthma). Children who were EBV
seropositive at 2 years of age were less likely to exhibit IgE sensitization to food and
inhalant allergens (88,89). A similar inverse association between EBV seropositivity and
sensitization was observed in older children (90). In another study, children infected with
at least one herpesvirus (HHV1, HHV2, HHV3, HHV6) by the age of 3 were less likely
to develop asthma by age 7 (91).
SPECIFIC AIMS
The specific sims for each study in this dissertation are described below:
Study 1: History of infectious mononucleosis and immune conditions: a population-
based cross-sectional study among twins
Aim 1. Determine whether self-reported history of infectious mononucleosis is associated
with self-reported history of atopic conditions in a population-based sample of twin pairs.
Aim 2. Determine whether self-reported history of infectious mononucleosis is associated
with self-reported history of autoimmune disease in a population-based sample of twin
pairs.

16
Study 2: Infectious mononucleosis, immune genotypes, and non-Hodgkin lymphoma
(NHL): a pooled case-control study from the InterLymph Consortium
Aim 1: Examine the joint effects of infectious mononucleosis history and a candidate
panel of 12 immune-related genetic variants on the risk of non-Hodgkin lymphoma.
Aim 2: Explore the impact of infectious mononucleosis history on age at non-Hodgkin
lymphoma diagnosis.
Aim 3: Investigate associations between infectious mononucleosis, demographics, and
family structure among controls.
Aim 4: Investigate associations between infectious mononucleosis and a candidate panel
of 12 immune-related genetic variants among controls.
Study 3: A genome-wide association study of Epstein-Barr virus viral copy number
and antibodies to Epstein-Barr virus antigens among Hodgkin Lymphoma survivors
Aim 1: Determine whether genetic variation is associated with Epstein-Barr virus viral
copy number among Hodgkin lymphoma survivors.
Aim 2: Determine whether genetic variation is associated with the presence of antibodies
to Epstein-Barr virus antigens among Hodgkin lymphoma survivors.
Aim 3: Determine whether genetic variation is associated with the levels of antibodies to
Epstein-Barr virus antigens among Hodgkin lymphoma survivors.

17
TABLES
Table 1.1. Known or purported function of EBV Latent Antigens. Adapted from
Macsween and Crawford (2003) and Sullivan (2019).
EBV latent antigen Function
EBNA-1 Episome maintenance
EBNA-2 Viral oncogene, transactivates cellular and other latent viral genes
EBNA-3A Viral oncogene, activates cellular genes
EBNA-3B Activates cellular genes
EBNA-3C Viral oncogene, increases LMP-1 expression
EBNA-LP Co-activates EBNA-2 responsive genes, increases efficiency of immortalization, RNA processing
LMP-1 Viral oncogene, induces B-cell activation and adhesion, protects from apoptosis
LMP-2 Repression of lytic cycle, enhances B-cell survival
EBNA: Epstein-Barr virus nuclear antigen; EBV: Epstein-Barr virus; LMP: latent membrane protein; LP: leader protein.

18
Table 1.2. EBV-Associated Lymphoid Malignancies in Nonimmunocompromised Persons.
Adapted from Fields Virology 2013.
Disease EBV frequency
EBV gene
expression
EBV
latency Cell of origin
Latency period
after EBV
infection
Burkitt lymphoma Endemic: 95–100%;
Sporadic: 15–85%
EBNA-1 1 B-cell GC
centroblast
3–8 years
Hodgkin lymphoma LD: 80–90%;
MC: 60–100%;
NS: 20–40%
EBNA-1,
LMP1,
LMP2
2 B-cell post-GC 1 year or more
Diffuse large B-cell
lymphoma, NOS
Up to 25% EBNA-1,
LMP1,
LMP2, and/or
EBNA-2
2 or 3 B-cell, GC, or
post-GC
>30 years
Peripheral T-cell
lymphoma
associated with
chronic acute or
acute EBV infection

100% EBNA-1,
LMP1, and/or
LMP2
2 Mature CD4+ T-
cell
None or several
months
Angioimmunoblastic
T-cell lymphoma
>90% EBNA-1,
LMP1,
LMP2
2 Lymphoma in
CD4 T-cells; EBV
in B-cells, GC
>30 years
Pyothorax lymphoma 100% EBNA-1,
EBNA-2,
and/or
LMP1
3 B-cell, GC, or
post-GC
>30 years
Extranodal NK/T
nasal
100% EBNA-1,
LMP1, and/or
LMP2
2 CD3−, CD56+ >30 years
EBNA: Epstein-Barr virus nuclear antigen; EBV: Epstein-Barr virus; GC: germinal center; LD: lymphocyte depleted;
LMP: latent membrane protein; MC: mixed cellularity; NS: nodular sclerosis; NOS: not otherwise specified.

19
Table 1.3. EBV-Associated Lymphoid Malignancies in Immunocompromised Persons.
Adapted from Longnecker, Kieff, Cohen (2013).
Affected persons Disease
EBV
frequency
EBV gene
expression
EBV
latency
Cell of
origin
Latency
period after
EBV infection
Patients with
congenital or
iatrogenic T-cell
immunodeficiencies
Lymphoprolife
rative disease
in immune
deficiency
100% EBNA-1,
EBNA-2,
EBNA-3,
EBNA-LP,
LMP1,
LMP2
3 B-cell, GC,
or post-GC
<3 months
Immunosuppressed
transplant recipients
Posttransplant
lymphoprolifer
ative disease
Early: <1 year
after
transplant
>90%;
Late: EBV
less common
EBNA-1,
EBNA-2,
EBNA-3,
EBNA-LP,
LMP1,
LMP2
3 B-cell, GC,
or post-GC
Early: <1 year
Late: >1 year
HIV patients Burkitt
lymphoma
30–70% EBNA-1 1 B-cell, GC Early in HIV

Hodgkin
lymphoma
>95% EBNA-1,
LMP1,
LMP2
2 B-cell, post-
GC
Early in HIV

DLBCL-
centroblast

30% EBNA-1 1 B-cell, GC Early in HIV

DLBCL-
immunoblastic
>90% EBNA-1,
EBNA-2,
EBNA-3,
EBNA-LP,
LMP1,
LMP2
3 B-cell, GC,
or post-GC
Late in HIV

Primary CNS
lymphoma
>95% EBNA-1,
EBNA-2,
EBNA-3,
EBNA-LP,
LMP1,
LMP2
3 B-cell, GC,
or post GC
Late in HIV

Primary
effusion
lymphoma
>90% EBNA-1,
and/or
LMP1,
LMP2

1 or 2 B-cell, GC,
or post-GC
Late in HIV

Plasmablastic
lymphoma
50%–80% EBNA-1 1 B-cell, post-
GC
Late in HIV

Smooth
muscle tumors
>95% EBNA-2
and/or
LMP1
? Smooth
muscle cell
Late in HIV
CNS: central nervous system; DLBCL: diffuse large B-cell lymphoma; EBV: Epstein-Barr virus; EBNA: Epstein-Barr
virus nuclear antigen; GC: germinal center; HIV: human immunodeficiency virus; LMP: latent membrane protein; LP:
leader protein.

20
Chapter 2. HISTORY OF INFECTIOUS MONONUCLEOSIS AND IMMUNE
CONDITIONS: A POPULATION-BASED CROSS-SECTIONAL STUDY AMONG
TWINS
ABSTRACT
Infectious mononucleosis (IM), a clinical syndrome typified by fever, pharyngitis,
and lymphadenopathy occurs in a subset of adolescents and young adults who experience
delayed primary Epstein-Barr virus infection. We previously reported higher
monozygotic concordance for IM among a large cohort of California twins. Here, we
explored associations between IM and other immune-related conditions using the same
population to determine whether there was an underlying shared predisposition.
Participant data were abstracted from the self-completed California Twin Program
enrollment questionnaire. Associations between IM and immune conditions (atopic,
autoimmune) were evaluated among like-sex pairs using unadjusted conditional logistic
regression models matched on twin pair (all race/ethnic groups) and multi-level logistic
regression models adjusted for twin pair, response age, highest parental education level,
and healthcare seeking behavior. Multi-level models were restricted to white/European
American (EA) twins and stratified by sex. Among 3699 EA twin pairs included in multi-
level analyses, the odds of IM were elevated in twins reporting at least one atopic
condition (OR = 1.71. 95% CI: 1.36, 2.16) or allergic rhinitis (OR = 1.58. 95% CI: 1.24,
2.01) compared to those who did not. Evidence of positive associations between IM and
both allergic rhinitis and other atopic conditions warrants further exploration.

21
Infectious mononucleosis (IM), the clinically apparent response to primary
Epstein-Barr virus (EBV) (and rarely cytomegalovirus [CMV]) infection, is characterized
by tonsillar pharyngitis, lymphadenopathy, and fever. It has an approximate incidence of
500 cases per 100,000 person-years in the United States (3). In industrialized countries
and among populations of high socioeconomic status, many individuals escape primary
EBV infection in childhood, which is typically asymptomatic or mild (6,46,92), and are
exposed to EBV in adolescence or young adulthood. IM occurs in a subset of those
individuals (93). The severity of primary EBV infection ranges from asymptomatic
seroconversion to nonspecific subclinical febrile illness to full-blown classical IM with
potentially serious sequelae. The risk and severity of symptomatic primary infection
increase with age (4).
Because only a proportion of individuals with delayed EBV exposure develop IM,
genetic susceptibility has been proposed. In support of this hypothesis, we observed
increased IM concordance among monozygotic compared to dizygotic twins enrolled in
the California Twin Program (CTP) (32). The IM concordance ratio among monozygotic
twin pairs was roughly double the concordance ratio of dizygotic twin pairs. Among
unaffected like-sex co-twins of IM cases, those from monozygotic twin pairs were 2.5
times as likely to develop IM (95% CI: 1.2, 5.3) (32). This finding was replicated in a
registry-based study in Denmark and was extended to include other types of familial
aggregation of IM (33).
Associations have been reported between IM (development and severity) and
polymorphisms in the human leukocyte antigen (HLA) class I and II regions, which may

22
affect immune response by altering the efficiency of presentation of viral peptides to T-
cells (34–36). Associations between polymorphisms in the IL10 gene, which codes for an
anti-inflammatory cytokine by the same name, and severity/age of EBV seroconversion
have been reported as well (37,38).
A strong association between IM and multiple sclerosis, an inflammatory
demyelinating neurological disease, was reported in a meta-analysis (10). While EBV has
been implicated as a possible etiologic agent for systemic lupus erythematosus (9) and
rheumatoid arthritis (81), associations between these autoimmune conditions and IM
have not been studied extensively.
Several studies have also reported associations between age at EBV infection and
atopic conditions (allergies, chronic eczema, and asthma). Children who were EBV
seropositive at 2 years of age were less likely to exhibit Immunoglobulin E (IgE)
sensitization to food and inhalant allergens (88,89). A similar inverse association between
EBV seropositivity and sensitization was also observed in children 4–13 years of age
(90). In another study, children infected with at least one herpesvirus (HHV1, HHV2,
HHV3, HHV6) by the age of 3 were less likely to develop asthma by age 7 (91).
Early childhood environment is an important predictor for age at primary EBV
infection and risk of IM (46) as well as atopic conditions. Because twins are well-
matched on early childhood environment, genome (monozygotic twins), and childhood
socioeconomic status, they provide a unique opportunity to study associations while
controlling for a variety of known and unknown genetic and environmental factors.

23
The purpose of this analysis was to explore potential associations between
immune-related (atopic and autoimmune) medical conditions and IM among twins. We
hypothesized that a positive IM history would be positively associated with both atopic
and autoimmune medical conditions in our study population.

MATERIALS AND METHODS
The University of Southern California Institutional Review Board granted
approval for this study, and each participant provided signed informed consent according
to the WMA Declaration of Helsinki Ethical Principles for Medical Research Involving
Human Subjects in 1964.
Study Population
Data for this study were collected as part of the CTP, a population-based cohort of
51,609 twins born in California between 1908 and 1982 with verifiable addresses who
opted into the cohort by completing a study questionnaire. Additional details about study
design, including ascertainment and recruitment of participants, have been described
elsewhere (94).
The primary analyses in this study are restricted to twin pairs born between 1957
and 1982 who completed an updated questionnaire including detailed questions about
medical history. Both monozygotic (MZ) and dizygotic (DZ) twin pairs were eligible for
inclusion. Twins who were adopted and those with unknown zygosity were excluded as
were MZ twin pairs who reported discordant genders. Female-male twin pairs were also

24
excluded due to potential confounding by sex. Single-respondent twin pairs were
excluded due to low sensitivity of proxy-reported IM history.
Data collection
Participant demographic data (race/ethnicity, age, gender, zygosity) and self-
reported medical history (IM, autoimmune conditions [Crohn’s disease, dermatomyositis,
Grave’s disease, Hashimoto’s thyroiditis, lupus erythematosus, multiple sclerosis,
myasthenia gravis, primary biliary cholangitis/cirrhosis, psoriasis, rheumatoid arthritis,
scleroderma, ulcerative colitis] and atopic conditions [asthma, allergic rhinitis, chronic
eczema, animal allergies, and plant allergies]) were abstracted from a self-completed 16-
page questionnaire distributed to twins upon enrollment in the CTP.
Medical history data from double-respondent twin pairs were used to calculate
agreement between self-reported data and proxy-reported data collected from co-twins.
Statistical Analysis
Evaluation of co-twin proxy medical history: To determine the degree of
information bias we might introduce by including data from single-respondent twin pairs,
we evaluated the sensitivity and specificity of co-twin reported medical history as a proxy
for self-reported medical history. We also evaluated whether proxy-reported
misclassification of IM history was associated with proxy-reported misclassification of
other medical conditions. Data from double-respondent pairs were used for these
calculations.
Logistic regression models: Two statistical approaches were used to determine
associations between IM and other medical conditions: [1] conditional logistic regression

25
models matched on twin pair (matched) and [2] multi-level logistic regression models
using individual-level and twin pair-level data (multi-level). All models were restricted to
like-sex pairs. Due to low numbers of IM discordant twin pairs, matched models were
unadjusted. Multi-level models were stratified by sex and included a random intercept for
twin pair and fixed effects for response age, highest parental education level (0–11, 12,
13–15, 16+, or unknown years of schooling), and healthcare seeking behavior
(comprehensive physical exam and/or preventive care screening within the last year).
Due to the paucity of data available for other race/ethnic groups, multi-level models were
restricted to participants who self-identified as white/European American (EA).
Sensitivity analysis: To account for differences between MZ and DZ twin pairs,
we tested for effect modification by zygosity. Multi-level models were evaluated for
possible confounding by fixed effects for birth order, number of additional siblings,
twins’ education level, and advanced maternal age at twins’ birth (35 years of age or
older).
All statistical tests were two-sided, and the significance level was set at α = 0.05.
Conservative Bonferroni corrections were used to account for multiple comparisons. An
odds ratio (OR), 95% confidence interval (CI), P value, and Bonferroni-adjusted P value
(PBon) were computed for each logistic regression model. Statistical analysis was
performed using R (Vienna, Austria) (95) using the lme4 (96), survival (97), and MASS
(98) packages.

26
RESULTS
Co-twin proxy medical history
5,325 double-respondent twin pairs (10,650 individual twin respondents) who met
inclusion criteria were used to evaluate proxy-reported medical history. We observed low
sensitivity of proxy-reported IM (39%) and atopic conditions (60%). Among proxy
reports, IM status was differentially misclassified with respect to atopic condition status;
twins who self-reported an atopic history were 29% more likely to have their IM status
misclassified by co-twin proxy (P = 0.002). Analogously, twins who self-reported a
positive IM history were 71% more likely to have their atopic condition history
misclassified by co-twin proxy (P < 0.001). Lastly, we observed an association between
IM history misclassification and atopic condition history (P < 0.001). Because of the
differences observed between self- and proxy-reported medical history for variables of
interest, we limited all analysis to self-reported medical histories from double-respondent
twin pairs.
Participant characteristics
TWIN PAIR CHARACTERISTICS (ZYGOSITY, SEX, RACE/ETHNICITY, MOTHER’S
AGE AT TWINS’ BIRTH, NUMBER OF SIBLINGS, BIRTH ORDER, RESPONSE AGE,
AND IM STATUS) ARE SHOWN IN TABLES
Table 2.1. Most twin pairs (90%) did not report a history of IM. Female-female
twin pairs and MZ twin pairs accounted for 68% and 55% of the included pairs,
respectively. Respondents self-identified into eight mutually exclusive race/ethnic

27
groups. EA (74%); Latino, Hispanic, or Mexican-American (11%); and multiple (7%)
were the most frequently reported race/ethnic groups. 10% of twins were born to mothers
of advanced maternal age (35+ years of age), and 30% of twins had no older maternal
siblings (mother’s first pregnancy), and the majority (73%) reported at least one
additional sibling.
Compared to IM-negative twin pairs, IM-concordant twin pairs were more likely
to be EA (96% vs 72%, P < 0.001), MZ (75% vs 54%, P = 0.005), and born to families in
which at least one parent completed 13 or more years of schooling (82% vs 57%, P <
0.001). Female sex (P = 0.44), additional siblings (P = 0.41), and mother’s advanced
maternal age at twins’ birth (P = 0.41) were not more prevalent in IM-concordant
compared to IM-negative pairs.
Matched analysis among discordant twin pairs
Of the 5325 twin pairs that met inclusion criteria for this study, 465 were
discordant for IM. We did not observe statistically significant associations between IM
and any atopic or autoimmune conditions in matched analyses (Table 2.2).
Multi-level analysis
Of the 3946 EA twin pairs in our study, we were unable to calculate an age at
questionnaire response for 297 pairs (missing birth date and/or date of questionnaire
response). Multi-level models of associations between IM history and history of atopic
medical conditions or autoimmune disease among the remaining 3699 twin pairs are
described in Table 2.3. We observed strong evidence of an association between IM status
and a combined group of atopic conditions (asthma, allergic rhinitis, chronic eczema,

28
animal allergies, and plant allergies). After accounting for a random effect of twin pair
and fixed effects for age at survey response, parental education level, and healthcare
seeking behavior, the odds of IM were 1.71 times as high in twins who reported an atopic
condition compared those who did not (95% CI: 1.36, 2.16). This association withstood a
conservative Bonferroni adjustment for multiple comparisons (PBon < 0.001).
Directionally similar effects were observed for specific atopic conditions. The
odds of IM were 1.58 times as high in twins who reported allergic rhinitis compared
those who did not (95% CI: 1.24, 2.01, PBon = 0.001). After stratifying by sex,
associations between IM and both allergic rhinitis (OR = 1.52. 95% CI: 1.15, 2.02. PBon =
0.02) and the combined group of atopic conditions (OR = 1.71. 95% CI: 1.30, 2.24. PBon
<0.001) were observed among female twins. There was evidence of similar associations
among male twins, but the associations did not withstand adjustment for multiple
comparisons (PBon > 0.05). We did not observe strong evidence of an association between
IM and other atopic conditions (P > 0.05).
Although we observed an association of similar magnitude for a combined group
of autoimmune diseases (OR = 1.51, 95% CI: 0.94, 2.42), the relationship did not
approach statistical significance after accounting for multiple comparisons (PBon = 0.68).
Statistical power was insufficient to compute effect estimates for [1] the presence of at
least one autoimmune disease among male twins and [2] specific autoimmune diseases.
Sensitivity analysis
Although we made an a priori decision to stratify all multi-level results by sex,
we did not observe effect modification by sex on the association between IM and atopy

29
or between IM and allergic rhinitis in multi-level models (Pinteraction = 0.87 and 0.82,
respectively). Similarly, we did not observe effect modification by zygosity on the
associations between IM and atopy or allergic rhinitis (Pinteraction = 0.84 and 0.38,
respectively). Lastly, there was no evidence of confounding by fixed effects for advanced
maternal age at twins’ birth (35 years of age or older), birth order, sibship size, or twins’
education level (<5% change in OR).
DISCUSSION
Among the 3699 EA twin pairs included in our multi-level analyses, we observed
strong evidence of an association between IM and a combined group of atopic conditions.
We also observed an association of similar magnitude between autoimmune conditions
and IM, but the relationship did not approach statistical significance after accounting for
multiple comparisons.
IM and atopy both represent a deviation from the typical T-cell response to
common immunological stimuli. It has been hypothesized that the severity of primary
EBV infection and the development of IM are attributable, at least in part, to an amplified
EBV-specific CD8+ and CD4+ T-cell response which is not observed in those whose
EBV seroconversion occurs asymptomatically (20,25,31). Atopy is a genetic
predisposition to produce IgE antibodies in reaction to allergen exposure due to a T
helper type 2 (TH2) response from CD4+ T helper cells (82). The observed association
between IM and atopy may result from a combination of shared environmental risk
factors, genetic susceptibility, or underlying immune abnormality.

30
Primary EBV infection during childhood—and subsequent avoidance of IM—is
dependent on exposure to EBV via saliva from infected individuals, and cultural norms
related to exposure to microbes appear to be associated with early EBV infection (99).
Similarly, increased childhood exposure to microbes may attenuate risk of atopic
conditions. For example, children whose parents used sucking as a method of pacifier
cleaning were less likely to develop asthma and eczema than children whose parents used
other cleaning methods (100). A similar inverse association has been observed between
atopic conditions and hand dishwashing (101). Taken together, these findings indicate
lenient parental attitudes toward microbe exposure may protect against both IM and
asthma.
IM heritability has been documented by our group and others (32,33). The
heritability of atopic conditions is well established and has been reported extensively
(102). A growing body of evidence suggests association between IM and variants in
the human leukocyte antigen (HLA) class I and II regions (34–36). Associations between
HLA polymorphisms have also been reported for atopic conditions in diverse populations
(103–105). Although no overlapping risk variants in HLA regions have been reported for
atopy and IM, the possibility of shared genetic risk factors, particularly those related to
the major histocompatibility complex, may explain the observed association between IM
and atopic conditions.
The relationship between IM and atopy may be confounded by participants’
likelihood to seek medical care. In order to approximate the degree of confounding
attributable to healthcare seeking behavior, we dichotomized participants based on

31
whether or not they had received a comprehensive physical exam or preventive screening
within the last year before completing the study questionnaire. We observed a 1% change
in the strength of the association between IM and atopy after adding a fixed effect for
healthcare seeking behavior to the multi-level logistic regression model, indicating a
minimal confounding effect of adult propensity for seeking medical care. Healthcare
seeking behavior among adult participants is an imperfect proxy for tendency to seek
medical care during childhood and adolescence when atopy and IM are more likely to
present. We accounted for childhood socioeconomic status (approximated by parental
educational attainment) in our multi-level models, which may also capture familial
tendency for seeking medical care.
It is possible the self-reported medical history used for this study introduced recall
bias. IM is a serious medical event of significant duration, and participants are unlikely to
have misremembered an IM diagnosis. It is also possible some of the correctly recalled
diagnoses may not have been serologically confirmed by the diagnosing physician.
However, symptomatic severity of atopic conditions can vary widely (86,106), and it is
likely some percentage of participants incorrectly classified themselves. Nevertheless,
participants were unaware of the study hypothesis when they completed the 16-page
questionnaire and answered questions about several medical conditions. Thus,
misclassification of atopy is likely to be nondifferential with respect to IM resulting in
bias toward the null.
Although we observed a strong association between IM and atopic conditions in
our multi-level model, we did not observe similar evidence in our analysis of IM

32
discordant twin pairs, which are matched on early childhood environment and at least
partially on genetics. It is unlikely this discrepancy between results is attributable to a
lack of statistical power. A power calculation using McNemar’s test for the matched
design indicated 5325 twin pairs provided >99% power to detect an OR of 1.71 with α =
0.05 and an 8.7% proportion of IM-discordant pairs. Instead, IM and atopy might both be
partially attributable to childhood exposures which are commonly shared by twins (both
MZ and DZ), thereby rendering the association between IM and atopy undetectable in the
matched design. In our multi-level analysis, we did not observe confounding by sibship
size or twins’ birth order relative to other siblings, which are commonly associated with
childhood exposure to infection. The majority of twins in our sample came from families
with less than five children thereby decreasing our ability to study the effect of large
sibship sizes. A more nuanced approach to gauging childhood environmental influence
on immunological development is warranted but was not possible given the limitations of
available study data.
Because our study population of twins comprises a population-based sample of
Californians, the results from our multi-level models are likely to be generalizable to
similar populations of European ethnicity. However, these results should not be
generalized to populations of other ethnicities without further exploration via targeted
studies.
In a large cohort of California born twins, we previously showed that IM is
heritable (32). Here we observed strong evidence of positive associations between IM
and both allergic rhinitis and a combined group of atopic conditions among EA twins.

33
Our study was underpowered to detect associations between IM and specific autoimmune
conditions. We recommend further exploration of these findings to investigate possible
common genetic risk factors between atopy and IM, and to employ more reliable methods
of ascertaining medical history. Studying this association in broader population groups is
also necessary.

34
TABLES
Table 2.1. Characteristics of twin pairs by IM status
IM
Negative

IM
Discordant

IM
Concordant

N=4811 pairs

N=465 pairs

N=49 pairs
N (%)

N (%)

N (%)
Zygosity and sex

Monozygotic Female-Female 1747 (36%)

185 (40%)

27 (55%)

Monozygotic Male-Male 877 (18%)

60 (13%)

10 (20%)

Dizygotic Female-Female 1486 (31%)

167 (36%)

9 (18%)

Dizygotic Male-Male 701 (15%)

53 (11%)

3 (6%)

Race/Ethnicity

Black or African American 136 (3%)

5 (1%)

0 (0%)

Chinese, Japanese, or Korean 149 (3%)

4 (1%)

0 (0%)
Filipino, Vietnamese, or South East Asian 42 (1%)

1 (0%)

0 (0%)

Latino, Hispanic, or Mexican 573 (12%)

12 (3%)

1 (2%)
Native American or American Indian 31 (1%)

1 (0%)

0 (0%)

White or European 3486 (72%)

413 (89%)

47 (96%)

Multiple 339 (7%)

23 (5%)

1 (2%)

Other 44 (1%)

4 (1%)

0 (0%)

No response 11 (0.2%)

2 (0%)

0 (0%)

Mother's age at twin pair's birth (years)

≤ 34 years of age 3571 (74%)

365 (78%)

40 (82%)

35 years of age or older 502 (10%)

43 (9%)

3 (6%)

No response 738 (15%)

57 (12%)

6 (12%)

Number of additional siblings

0 621 (13%)

68 (15%)

9 (18%)

1 1314 (27%)

165 (35%)

13 (27%)

2 or more 2201 (46%)

179 (38%)

23 (47%)

No response 675 (14%)

53 (11%)

4 (8%)

Birth order

1st pregnancy 1464 (30%)

140 (30%)

16 (33%)

2nd pregnancy 1411 (29%)

159 (34%)

15 (31%)

3rd or later pregnancy 1602 (33%)

127 (27%)

13 (27%)

No response 334 (7%)

39 (8%)

5 (10%)

Maximum years of parental schooling (mother or father)

0–11 498 (10%)

23 (5%)

1 (2%)

12 1175 (24%)

89 (19%)

7 (14%)

13–15 1056 (22%)

102 (22%)

15 (31%)

16+ 1678 (35%)

213 (46%)

25 (51%)

Mean (SD) Mean (SD) Mean (SD)
Age at questionnaire response
a
34.1 (5.6) 34.9 (4.9) 34.4 (4.9)
Abbreviations: IM, infectious mononucleosis; SD, Standard deviation.
a
Based on the data provided, age at questionnaire response could not be calculated for 31% of twins (N = 3252
participants)

35
Table 2.2. Risk of IM according to history of atopic or autoimmune medical conditions
among 465 IM-discordant like-sex twin pairs
N Twin Pairs
P value

IM case
exposed/
IM case
unexposed/

Medical condition
Co-twin
unexposed
Co-twin
exposed OR
a
95% CI
Allergic rhinitis 87 70 1.24 0.91, 1.7 0.18
Animal or plant allergy 59 69 0.86 0.6, 1.21 0.38
Asthma 40 49 0.82 0.54, 1.24 0.34
Chronic eczema 20 17 1.18 0.62, 2.25 0.62
Any atopic condition 90 79 1.14 0.84, 1.54 0.40

Chron's disease 1 2
--
b

Dermatomyositis 0 0
--
b

Grave’s disease 4 0
--
b

Hashimoto’s thyroiditis 3 0

Lupus erythematosus 3 2 1.50 0.25, 8.98 0.66
Multiple sclerosis 1 1
--
b

Myasthenia gravis 1 0
--
b

Primary biliary cholangitis 0 1
--
b

Psoriasis 6 6 1.00 0.32, 3.1 >0.99
Rheumatoid Arthritis 6 9 0.67 0.24, 1.87 0.44
Scleroderma 0 1
--
b

Ulcerative colitis 7 4 1.75 0.51, 5.98 0.37
Any autoimmune condition 29 24 1.21 0.7, 2.08 0.49
Abbreviations: IM, infectious mononucleosis; OR, odds ratio; CI, confidence interval.
a
ORs, 95% CIs, and P values calculated using conditional logistic regression models matched on twin pair
b
Insufficient data to compute effect estimate.

36

Table 2.3. Infectious mononucleosis association with atopic conditions and autoimmune disease among 3699 like-sex white/European
American twin pairs
Medical condition(s)
a

Condition not reported: N (%)

Condition reported: N (%)
Gender IM-negative IM-positive

IM-negative IM-positive OR
b
95% CI P value
Allergic rhinitis Female 3257 (95%) 181 (5%)

1432 (92%) 122 (8%) 1.52 1.15, 2.02 0.003 **
Male 1738 (96%) 71 (4%)

560 (94%) 37 (6%) 1.64 1.01, 2.68 0.05

All 4995 (95%) 252 (5%)

1992 (93%) 159 (7%) 1.58 1.24, 2.01 < 0.001 **

Animal or plant allergy Female 3838 (94%) 236 (6%)

851 (93%) 67 (7%) 1.16 0.83, 1.61 0.39

Male 1939 (96%) 83 (4%)

359 (93%) 25 (7%) 1.61 0.93, 2.8 0.09

All 5777 (95%) 319 (5%)

1210 (93%) 92 (7%) 1.27 0.96, 1.69 0.10

Asthma Female 4098 (94%) 256 (6%)

591 (93%) 47 (7%) 1.23 0.84, 1.81 0.28

Male 2052 (96%) 92 (4%)

246 (94%) 16 (6%) 1.46 0.76, 2.8 0.26

All 6150 (95%) 348 (5%)

837 (93%) 63 (7%) 1.30 0.94, 1.81 0.12

Chronic eczema Female 4508 (94%) 289 (6%)

181 (93%) 14 (7%) 1.21 0.63, 2.33 0.56

Male 2270 (96%) 104 (4%)

28 (88%) 4 (13%) --
c

All 6778 (95%) 393 (5%)

209 (92%) 18 (8%) 1.49 0.83, 2.68 0.18

Any atopic condition Female 2697 (95%) 134 (5%)

1992 (92%) 169 (8%) 1.71 1.3, 2.24 < 0.001 **
Male 1464 (96%) 55 (4%)

834 (94%) 53 (6%) 1.67 1.06, 2.63 0.03 *
All 4161 (96%) 189 (4%)

2826 (93%) 222 (7%) 1.71 1.36, 2.16 < 0.001 **

Any autoimmune condition Female 4359 (94%) 273 (6%)

330 (92%) 30 (8%) 1.51 0.94, 2.42 0.09

Male 2173 (96%) 102 (4%)

125 (95%) 6 (5%) --
c

All 6532 (95%) 375 (5%)

455 (93%) 36 (7%) 1.41 0.92, 2.14 0.11

Abbreviations: CI, confidence interval; IM, infectious mononucleosis; OR, odds ratio.

* P value < 0.05 before adjusting for multiple comparisons, ** P value < 0.05 after Bonferroni adjustment for multiple comparisons.

a
Any atopic condition: allergic rhinitis, animal allergy, plant allergy, asthma, chronic eczema. Any autoimmune condition: Crohn’s disease, dermatomyositis, Grave’s disease,
Hashimoto’s thyroiditis, lupus erythematosus, multiple sclerosis, myasthenia gravis, primary biliary cholangitis/cirrhosis, psoriasis, rheumatoid arthritis, scleroderma, ulcerative
colitis. Data were insufficient to compute effect estimates for individual autoimmune conditions.

b
ORs, 95% CIs, and P values calculated using multi-level logistic regression models with a random intercept for twin pair and fixed effects for response age, highest parental
education level (0-11, 12, 13-15, or 16+ years of schooling), and advanced maternal age at twins’ birth (35 years of age or older). Twins missing response age were excluded from
analysis (N=494).

c
Insufficient data to compute effect estimate for chronic eczema and autoimmune conditions among male twin pairs.

37

Chapter 3. INFECTIOUS MONONUCLEOSIS, IMMUNE GENOTYPES, AND
NON-HODGKIN LYMPHOMA (NHL): A POOLED CASE-CONTROL STUDY
FROM THE INTERLYMPH CONSORTIUM

PUBLICATION AUTHORS
Niquelle Brown Wadé, Cindy M Chang, David Conti , Joshua Millstein , Christine
Skibola, Alexandra Nieters, Sophia S Wang, Silvia De Sanjose , Eleanor Kane, John J
Spinelli, Paige Bracci, Yawei Zhang, Susan Slager, Jun Wang , Henrik Hjalgrim, Karin
Ekstrom Smedby,

Elizabeth E Brown, Ruth F Jarrett, Wendy Cozen

38
AUTHOR AFFILIATIONS
Department of Preventive Medicine, Keck School of Medicine, University of
Southern California, Los Angeles, California, USA (Niquelle Brown Wadé, David Conti,
Joshua Millstein, Jun Wang, Wendy Cozen); Cigna Health and Life Insurance Company
(Cigna), Bloomfield, CT, USA (Niquelle Brown Wadé); Division of Population Health
Sciences, Center for Tobacco Products, Food and Drug Administration, Bethesda,
Maryland, USA (Cindy Chang); USC Norris Comprehensive Cancer Center, Keck
School of Medicine, University of Southern California, Los Angeles, California, USA
(David Conti, Jun Wang, Wendy Cozen); Department of Hematology and Medical
Oncology, Emory University School of Medicine, Atlanta, Georgia, USA (Christine
Skibola); Center for Chronic Immunodeficiency (CCI), University Medical Center
Freiburg, Germany, University of Freiburg, Freiburg, Germany (Alexandra Nieters);
Department of Computational and Quantitative Medicine, City of Hope Comprehensive
Cancer Center, Duarte, CA, USA (Sophia S Wang); Sexual and Reproductive Health,
PATH, Seattle, Washington, USA (Silvia De Sanjose); Centro de Investigación
Biomédica en Red: Epidemiología y Salud Pública (CIBERESP), Madrid, Spain (Silvia
De Sanjose); Department of Health Sciences, University of York, York, YO10 5DD,
United Kingdom (Eleanor Kane); Population Oncology, BC Cancer Agency; School of
Population and Public Health, University of British Columbia, Vancouver, British
Columbia, Canada (John J Spinelli); Department of Epidemiology and Biostatistics,
University of California at San Francisco, San Francisco, California, USA (Paige Bracci);
Department of Surgery, Yale School of Medicine and Yale School of Public Health, New

39
Haven, Connecticut, USA (Yawei Zhang); Department of Epidemiology, Mayo Clinic,
Rochester, Minnesota, USA (Susan Slager); Department of Epidemiology Research,
Statens Serum Institut, Copenhagen; Department of Haematology, Rigshospitalet,
Copenhagen, Denmark (Henrik Hjalgrim); Karolinska Instiutet, Karolinska University,
Stockholm, Sweden University Hospital, Sweden (Karin Ekstrom Smedby); Department
of Pathology and O’Neal Comprehensive Cancer Center of the University of Alabama at
Birmingham, Birmingham, Alabama, USA (Elizabeth E Brown); MRC-University of
Glasgow Centre for Virus Research, Glasgow, Scotland (Ruth F Jarrett); Department of
Pathology, Keck School of Medicine, University of Southern California, Los Angeles
California, USA (Wendy Cozen).

40
ABSTRACT
We explored interaction between non-Hodgkin lymphoma (NHL), infectious
mononucleosis (IM) history, and immune-related genotypes in a pooled case-control
analysis. 7926 NHL patients and 10018 controls from 12 studies (17–96 years of age,
recruited 1988–2008) were included. Self-reported IM history and genotypes were
provided by the InterLymph Data Coordinating Center. Effect estimates were derived
using multivariate logistic and linear regression, and interaction effects were estimated
using the empirical Bayes method. Bonferroni corrections and PACT were used to account
for multiple comparisons. There was evidence of interaction between IM history and two
variants on T-cell lymphoma (TCL) risk: rs1143627 in interleukin-1B (Pinteraction = 0.04,
ORinteraction = 0.09, 95% confidence interval [CI] = 0.01, 0.87) and rs1800797 in
interleukin-6 (Pinteraction = 0.03, ORinteraction = 0.08. 95% CI = 0.01, 0.80). Neither
interaction effect withstood adjustment for multiple comparisons. Among controls,
increasing socioeconomic status (OR = 1.69. 95% CI = 1.48, 1.93) and female sex (OR =
1.53. 95% CI = 1.26, 1.87) were positively associated with IM. Large sibship size (3+)
was inversely associated with IM among controls born before 1960 (OR<1960 = 0.40. 95%
CI = 0.24, 0.67), but not after. Genetic risk variants in IL1B and IL6 may affect the
association between IM and TCL. Risk factors for IM are consistent with lower Epstein-
Barr virus exposure in early life; the association with female sex is unexplained.

41
FUNDING
This work was supported by awards from National Cancer Institute/National
Institutes of Health (N01-CN-75014-20, P30CA014089, R01 CA186646, P30 CA13148,
R21 CA155951, U54 CA118948, CA45614, CA87014, CA104682 and CA154643);
Agència de Gestió d’Ajuts Universitaris i de Recerca (AGAUR), CERCA
Programme/Generalitat de Catalunya for institutional support (2017SGR1085); Spanish
Ministry of Economy and Competitiveness - Carlos III Institute of Health cofunded by
FEDER funds/European Regional Development Fund (ERDF) - a way to build Europe
(PI14/01219); Centro de Investigación Biomédica en Red: Epidemiología y Salud Pública
(CIBERESP, Spain); the Canadian Institutes for Health Research (CIHR); Canadian
Cancer Society; and Michael Smith Foundation for Health Research [British Columbia]).
The collection of cancer incidence data used in the UCSF study was supported by the
California Department of Public Health pursuant to California Health and Safety Code
Section 103885; Centers for Disease Control and Prevention’s (CDC) National Program
of Cancer Registries, under cooperative agreement 5NU58DP003862-04/DP003862; the
National Cancer Institute’s Surveillance, Epidemiology and End Results Program under
contract HHSN261201000140C awarded to the Cancer Prevention Institute of California,
contract HHSN261201000035C awarded to the University of Southern California (NHL
MultiCenter Case-Control Study site), and contract HHSN261201000034C awarded to
the Public Health Institute.

42
DISCLAIMERS
The ideas and opinions expressed herein are those of the author(s) and do not
necessarily reflect the opinions of the State of California, Department of Public Health,
the National Cancer Institute, National Institutes of Health, or the Centers for Disease
Control and Prevention or their Contractors and Subcontractors. The information in this
chapter is not a formal dissemination of information by the FDA and does not represent
agency position or policy. The contents are the responsibility of the authors alone. This
chapter was prepared while Dr. Cindy Chang was employed at the National Cancer
Institute.

43
Non-Hodgkin lymphoma (NHL) comprises a group of lymphoid malignancies
with distinct histopathologies and risk patterns (68) originating from B- (~80%) and T-
lymphocytes (~20%). Genetic or acquired immunodeficiency is the strongest risk factor,
but more subtle immune alterations may also play a role in pathogenesis (107). For
example, there is a strong positive association between NHL and autoimmune disease
(108,109) and an inverse association with atopy (110). In addition to evidence of
familiality for overall and subtype-specific NHL risk (111,112), variants in and near
genes related to innate and adaptive immunity (IL1RN, FCGR2A, TNFA, HLA Class I
and II) (113–115) have been implicated as potential risk factors.
Several infectious agents, including Epstein-Barr virus (EBV) (116), Hepatitis C
virus (117), and Helicobacter pylori (118), contribute to NHL etiology through various
mechanisms including direct transformation of lymphocytes, immunosuppression,
chronic B-cell activation, and innate immune stimulation (119). EBV, a ubiquitous
member of the human herpesvirus family, induces B-cell growth by expression of viral
proteins and non-coding RNAs (6). The viral DNA persists as an episome in the host
memory B-cell DNA after infection where it remains latent in the presence of a
competent cytotoxic T-cell response. When acquired early in life, primary EBV infection
is generally asymptomatic or causes a mild, non-specific, febrile illness (29). In
industrialized countries and populations of higher socioeconomic status (SES), primary
infection is often delayed until adolescence or young adulthood. From 25% to 74% of
those experiencing delayed primary infection develop infectious mononucleosis (IM), a
moderate to severe clinical syndrome characterized by fever, tonsillar pharyngitis, and

44
lymphadenopathy (3,4,39). The severity of primary EBV infection and the development
of IM are attributable, at least in part, to an amplified EBV-specific CD8+ and CD4+ T-
cell response which is not observed in those whose EBV seroconversion occurs
asymptomatically (20,25,31). Propensity to develop the syndrome is influenced by
genetic factors related to immune response (34,36).
In the largest pooled case-control study of NHL conducted to date from the
International Lymphoma Epidemiology Consortium (InterLymph), Becker et al. (2012)
observed a positive association between self-reported IM history and risk of all NHL (OR
= 1.26. 95% CI = 1.01, 1.57). When stratified by subtype, associations were observed
between IM and T-cell lymphoma (TCL) and a B-cell category combining chronic
lymphocytic leukemia (CLL), small lymphocytic lymphoma (SLL), prolymphocytic
lymphoma (PLL), and mantle cell lymphoma (MCL) (120).
Studies indicating familial IM clustering and higher concordance for IM risk
among monozygotic compared to dizygotic twin pairs suggest a role for genetic
susceptibility in IM etiology (32,33). However, little is known about the influence of
genetic factors on IM risk or their role in modifying the possible association between IM
and NHL. Many of the genetic risk loci identified for NHL and NHL subtypes in
previous InterLymph studies are in or near genes related to immune response that might
also influence the association between IM and NHL risk (113,114,121–125).
In this InterLymph study, we examined the joint effects of IM history and 12
candidate immune-related variants on the risk of NHL, the impact of IM history on age at

45
NHL diagnosis, and risk factors for IM among controls using previously collected
demographic and familial information and the same candidate panel of immune-related
genetic variants.
MATERIALS AND METHODS
Study population
Participants included NHL patients and controls contributed from case-control
studies at InterLymph Consortium member sites: British Columbia: BC (Canada);
EpiLymph-Spain; EpiLymph-France; EpiLymph-Germany; EpiLymph-Italy; EpiLymph-
Ireland; EpiLymph-Czech Republic; Mayo Clinic (USA); National Cancer Institute
Surveillance, Epidemiology, and End Results: NCI-SEER (USA); Scandinavian
Lymphoma Etiology: SCALE (Denmark, Sweden); University of California San
Francisco: UCSF (USA); Yale (USA). All 12 studies (from 10 countries) had approval
from their respective National or Institutional Review Boards, and participants provided
signed informed consent according to the WMA Declaration of Helsinki Ethical
Principles for Medical Research Involving Human Subjects in 1964.
Seven participating sites used population-based ascertainment (population-based
case-control studies: British Columbia Cancer Agency BC, [Canada]; Scandinavian
Lymphoma Etiology: SCALE [Denmark, Sweden]; University of California San
Francisco: UCSF [USA]; National Cancer Institute Surveillance, Epidemiology, and End
Results: NCI-SEER [USA]; EpiLymph-Germany; Yale [USA]). Cases from these sites
were ascertained from population-based cancer registries or national health systems, and

46
controls were recruited from the same source population as the cases (census or random
digit dialing rosters or from the same national health system clinic practice, respectively).
Five sites used clinic or hospital-based ascertainment (clinic-based case-control
studies: EpiLymph-Spain, EpiLymph-France, EpiLymph-Ireland, EpiLymph-Czech
Republic, Mayo Clinic). Patients at these sites were identified from clinic or hospital
records and controls were identified from other patients without cancer attending the
same clinics.
Studies were conducted during various time periods between 1988 and 2008, and
participants were 17–96 years of age at the time of ascertainment/recruitment. Summaries
of study details are provided in Table 3.1 and Table 3.2. Additional details are available
in previous InterLymph publications (109,110,115,120,122–136).
InterLymph Consortium member studies were selected for inclusion based on the
availability of self-reported IM history and candidate variant genotypes from at least 50%
of participants. Participants who had missing data for age at enrollment, sex, SES, or IM
history were excluded. Because the number of non-white participants in member studies
was small and would require stratification for genetic analyses, we limited the study to
white participants. Consistent with previous InterLymph analyses, participants who
reported IM diagnosis less than 2 years before NHL diagnosis were excluded (120).
Data collection
The InterLymph Data Coordinating Center (Mayo Clinic, Rochester, MN)
harmonized data submitted by each study site into a de-identified, pooled dataset for

47
analysis. Information on demographics, family structure (number of siblings and birth
order), and IM history was self-reported using questionnaires (68). Ethnicity/race was
available for eleven of the twelve study centers included in the analysis, with the
participants from most of these European, U.S., and Canadian studies being non-Hispanic
white. Participants with missing race/ethnicity were included from SCALE (N = 5683),
Mayo Clinic (N = 28), Yale (N = 3), NCI-SEER-Seattle and Iowa (N = 20) studies since
the majority of the population in these study areas is non-Hispanic white; otherwise those
with missing race were excluded. Socioeconomic status (SES) was categorized based on
years of education (low: 0-12 years, high school or less; medium: 13-15 years, some
college; high: 16+ years, college degree or more) or tertiles of the SES variable submitted
by each individual study center.
The pooled analysis used existing genotype data on variants selected a priori
based on results from previous functional analyses, association with NHL, or role in pro-
/anti-inflammatory pathways (113,114,121–123). The effects of these 12 genetic variants
located in or near nine immune-response genes were assessed: IL1A-889C>T
(rs1800587), IL1B–511C>T (rs16944), IL1B–31T>C (rs1143627), IL1RN–9589A>T
(rs454078), IL2–384T>G (rs2069762), IL6–174G>C (rs1800795), IL6–597G>A
(rs1800797), IL10–3575T>A (rs1800890), IL10–1082A>G (rs1800896), TNF–308G>A
(rs1800629), HLA class I C>A (rs6457327), and HLA class II T>G (rs10484561).
Genotyping was performed using either TaqMan (Applied Biosystems, Inc., Foster City,
California), Pyrosequencing (Qiagen NV, Hilden, Germany), or Illumina Goldengate
(Illumina, Inc., San Diego, California) genotyping assays. Additional technical details

48
about genotyping methods used in each contributing study are included in previous
publications (113,122,123,135,137).
All NHL diagnoses were confirmed by pathology report review, with the majority
re-reviewed by a hematopathologist, depending on the study. NHL subtypes were
classified according to the World Health Organization (WHO) classification in 2001 and
2008 (138–140) and include chronic lymphocytic leukemia/small lymphocytic lymphoma
(CLL/SLL: ICD-O-3 codes 9670, 9823), diffuse large B-cell lymphoma (DLBCL: 9679,
9680, 9684), follicular lymphoma (FL: 9690, 9691, 9695, 9698), mantle cell lymphoma
(MCL: 9673), T-cell lymphoma (TCL: 9702, 9705, 9708, 9709, 9714, 9716, 9717, 9718,
9719, 9729, 9827, 9834), and all NHL combined (defined by the above ICDO3 codes and
9671, 9675, 9687, 9689, 9699, 9700, 9701, 9728, 9826, 9832, 9833, 9591, and 9727).
Patients with AIDS-related lymphomas were excluded.
Statistical Analysis
Candidate variants in linkage disequilibrium (LD): SNP Annotation and Proxy
Search (SNAP) (141) was used to assess LD via correlations between all pairs of
candidate variants in the same gene.
Main effect NHL associations: Unconditional logistic regression was used to
estimate odds ratios (ORs) and 95% confidence intervals (CIs) for the association
between IM and NHL and for associations between candidate genetic variants and NHL.
Consistent with other InterLymph publications (122,123), all genetic variants were coded
as dichotomous variables assuming a dominant model (absence or presence of minor

49
allele). All models were adjusted for age at NHL diagnosis/enrollment, sex, study center,
and SES.
Gene-environment interaction in NHL risk: The effect of interaction between IM
and immune-related genotypes on NHL risk was assessed using the empirical Bayes
approach described by Mukherjee et al. (142). Sensitivity analyses were then performed
using unconditional logistic regression to test the association between IM and NHL
stratified by each candidate variant genotype. Models were adjusted for the covariates
listed above. Associations were examined for all NHL combined and by NHL subtype.
Association with age at NHL diagnosis: We performed a case-only analysis
stratified by NHL subtype to assess the association between IM and age at NHL
diagnosis using linear regression models adjusted for sex, study center, SES, and year of
birth (to account for a potential birth cohort effect).
Risk factors for IM among controls: We evaluated the associations between IM
and demographic factors (SES and sex), family structure (birth order and sibship size),
and genetic factors (candidate genetic variants mentioned above) using unconditional
multivariable logistic regression models adjusted for age and study center. Models
including SES, family structure, or genetic variants as the exposure variable of interest
were adjusted for sex. Models including sex or genetic factors as the exposure variable of
interest were further adjusted for SES. Analyses with family structure as the exposure
variable of interest were stratified by year of birth (1901–1960, 1961–1990 based on a
historical shift in attitudes and policies about preschool attendance in the 1960s (143))

50
and restricted to sites with year of birth data for controls (Mayo Clinic; British Columbia,
Canada; and all EpiLymph sites). All analyses of risk factors for IM were restricted to
controls.
Sensitivity analysis and multiple comparisons: All genetic data were assessed for
deviations from allele frequencies expected under Hardy-Weinberg equilibrium among
controls, and a sensitivity analysis was conducted in which we excluded study centers
from the analysis of the specific genetic variants for which within-center allele
frequencies were inconsistent with Hardy-Weinberg equilibrium at P < 0.05. Additional
sensitivity analyses were conducted excluding studies using clinic-based control
recruitment methods.
All statistical tests were two-sided. For genetic analyses, the PACT statistic was
used to account for multiple comparisons and correlated tests from variants within the
same region (144). For analyses of demographic and familial factors, Bonferroni
corrected P values were used to account for multiple comparisons. Uncorrected P values
are reported in tables. For those associations with uncorrected P values <0.05,
Bonferroni-corrected P values (PBon) or PACT statistics are noted in the text. Statistical
analysis was performed using Stata, version 13 (StataCorp, LP, College Station, TX).

51
RESULTS
Main NHL associations
Summaries of study details are provided in Table 3.1 and Table 3.2. A total of
7926 NHL patients and 10018 controls from 12 InterLymph studies met the inclusion
criteria. The distribution of NHL patients and controls by selected demographic and
clinical characteristics is shown in Table 3.3 and Table 3.4, respectively. The majority
(83%) of patients were diagnosed with mature B-cell lymphoma (Table 3.5); the
remainder were diagnosed with mature T-cell (6%), precursor cell (1%), and missing
subtype/not otherwise specified (NOS) lymphomas (10%).
Analysis with SNAP indicated candidate risk variants in IL1B (r
2
IL1B: rs16944,
rs1143627 = 0.96) and in IL6 (r
2
IL6: rs1800795, rs1800797 = 0.97) were in high LD, and candidate
variants in IL10 were in moderate LD (r
2
IL10: rs1800890, rs1800896 = 0.66).
After adjustment for multiple comparisons, we observed strong main effects for
associations between HLA variants and NHL (PACT<0.001 and PACT = 0.004) and an
IL1RN variant and NHL (PACT = 0.04) (Table 3.6). A history of IM was associated with
all NHL combined (PBon = 0.06) and strongly associated with CLL/SLL (PBon = 0.04) and
MCL (PBon = 0.01) (Table 3.7). The direction of the association between IM and NHL
risk was consistent when restricted to using population-based methods for control
recruitment (not shown in tables). Thus, the main effects of genotype and IM for
associations with all NHL and NHL subtypes were largely consistent with previously
reported results from a subset of the same InterLymph studies (113,114,120–124).

52
Gene-environment interaction in NHL risk
Statistical interaction between candidate genetic variants and IM on risk of
CLL/SLL, DLBCL, FL, MCL, TCL, or all NHL combined are described in Table 3.8 and
Table 3.9). There was an interaction effect between a genetic variant in the IL1B gene
(rs1143627C) and IM history on TCL risk (ORinteraction = 0.09, 95% CI = 0.01, 0.87, p =
0.04) risk. We also observed interaction between rs1800797A in the IL6 gene and IM on
TCL risk (ORinteraction = 0.08, 95% CI = 0.01, 0.80, p = 0.04). Neither of these associations
persisted after adjustment for multiple comparisons (PACT > 0.05). These results were
directionally consistent when restricted to population-based case-control studies (not
shown in tables). Associations between IM history and TCL, stratified by IL1B and IL6
genotypes, are shown in Table 3.10. For each IL1B or IL6 variant, participants with the
minor allele have a lower risk of NHL. However, effect estimates are unstable due to low
sample sizes in strata comprised of IM-positive TCL patients. No interaction was
observed between other candidate variants and NHL or NHL subtypes (p ≥ 0.05).
Age at NHL diagnosis
Although self-reported history of IM was strongly associated with age at NHL
diagnosis among NHL patients of each subtype, the association was not present after
adjusting for birth year, suggesting a cohort effect (Table 3.11). We observed similar
results when the analysis was restricted to population-based studies.
IM risk factors among controls
Among 10018 control participants, 521 reported a positive history of IM.
Increasing SES level (ORtrend = 1.69. 95% CI = 1.48, 1.93. PBon < 0.001) and female sex

53
(OR = 1.53. 95% CI = 1.26, 1.87. PBon < 0.001) were positively associated with IM
(Table 3.12). These results were consistent when restricted to controls from population-
based case-control studies. We observed an association between sibship size and IM
history among all controls (Table 3.13, OR3+ siblings = 0.57. 95% CI = 0.38, 0.85).
However, stratification of controls by year of birth revealed a strong inverse association
between large sibship size (3+ siblings) and IM among those born before 1960 (PBon <
0.001). There was no evidence of an association between birth order and IM risk in
controls with 2 or more siblings (Table 3.13). None of the candidate variants showed
evidence of an association with IM history (Table 3.14).
DISCUSSION
IM was associated with an increased risk of TCL in the original main effects
InterLymph paper (120) and with a 32% (P = 0.17) increased risk among our subset of
InterLymph participants. The minor allele in variant rs1143627 in the promoter region of
the IL1B gene appeared to attenuate the effect of IM on TCL risk as did the minor allele
in variant rs1800797 in the promoter region of the IL6 gene, although the interaction
effects for both variants did not persist after adjustment for multiple comparisons.
IL1B, the cytokine encoded by the IL1B gene, is an inflammatory response and
fever mediator, and contributes to several lymphocyte activities including growth and
differentiation of B-cells (145), proliferation of T-helper Type 2 (TH2) clones (146), and
activation of Th17 cells (147). We observed a suggestive interaction effect of similar
magnitude between rs16944, an IL1B variant highly correlated with rs1143627, and TCL.
IL1B is required for T-cell activation in some immune responses (148,149) and thus

54
could contribute to increased T-cell replication. The minor alleles of the two IL1B
variants examined in our study are associated with lower expression of IL1B (150) and
may decrease T-cell activation in the setting of IM. This decrease in activation may, in
turn, attenuate the effects of the amplified T-cell response in IM. rs16944 has also been
associated with uncontrolled EBV replication in liver transplant patients, who later
develop post-transplant lymphoproliferative disorder (151), suggesting a link between IL-
1B and dysfunctional control of EBV. There was also suggestive association between the
functional variant rs1800797 in the IL6 gene promoter region and risk of TCL. Through
complex interactions with nearby variants, rs1800797 regulates the gene that encodes the
inflammatory cytokine IL6, which influences growth and differentiation of T-cells,
among many other immune functions (152,153).
In the presence of the significant T-cell expansion associated with IM, the
identified variants in IL1B and IL6 may reduce the chances of T-cell cell proliferation and
subsequent mutation or oncogenic rearrangement. These findings may extend to other
settings in which the T-cell compartment undergoes significant expansion, in particular,
during primary or reactivated viral infections. Follow-up of these observations in a
targeted study is warranted because of the potential biological pathway.
Among controls, increasing SES was associated with elevated risk of IM; the risk
of IM was roughly two times higher among high SES participants compared to low SES
participants. This observation is consistent with previous studies, which have suggested
high SES is a surrogate for a lower probability of EBV exposure in early life (due to
fewer siblings and less crowded environments) and thus, a higher risk of developing IM

55
(4,50,154,155). In our study, we observed a strong relationship between large sibship size
and IM among controls born before 1960 (PBon = 0.001) but not in those born after 1960
(P = 0.85). Controls born after 1960 may have been more likely to attend preschool,
which would provide EBV exposure in early life and simulate a large family.
Alternatively, it may be that after 1960, overcrowding decreased, and hygienic behaviors
generally increased, mitigating the importance of family size. The findings for our
controls born before 1960 are consistent with previous reports of inverse associations
between sibship size and IM (156,157). There was no evidence of an association between
birth order and IM history among controls with 2 or more siblings; however, the sample
size of controls born after 1960 with larger sibship size was insufficient to calculate
stable estimates of effect size because a number of study sites did not collect year of birth
data for controls.
In our study, IM was also more common among females (6% prevalence) than
males (5% prevalence), in both cases and controls. In a prospective study among
university students, Crawford et al. did not find a difference in IM prevalence among
EBV-seroconverters by gender (52); however higher rates of hospitalization for IM
among teen and young adult females in the UK were reported by Ramagopalan et al.
(158). While Crawford et al. measured propensity for IM upon primary EBV infection
(i.e. seroconverters) among college-age subjects, our study relied on self-reported IM
history among all subjects regardless of age, including IM that developed prior to and
after college. In addition, the denominator in the Crawford study included only EBV-
seronegative college students, while our denominator included adults of all ages, some of

56
whom surely acquired EBV prior to adolescence and were therefore not at risk of IM.
Thus, our results are not necessarily discrepant with those from the Crawford study.
Explanations for the higher IM prevalence among females in our study include lower
rates of EBV infection among females in childhood producing a higher pool at risk for
IM in adolescence and young adulthood, a higher reporting of symptomatic disease
(40,159,160), higher likelihood of medical care seeking behavior and thus diagnosis, or a
true biological effect enhancing IM risk in females compared to males at a wider range of
ages (young or older than college). Sex-based biological differences in response to
infections have been reported. For example, females mount more vigorous antibody- and
cell-mediated immune responses following some infections and vaccines than men
(161,162).
A limitation of our study is reliance on self-reported IM history, which could be
affected by recall bias. However, IM is a severe and debilitating syndrome of relatively
long duration, interrupting young adult life; therefore, it is unlikely that a participant
would forget this experience. Another limitation is the use of cross-sectional data for
determining IM risk factors which prevents the establishment of a temporal relationship
between IM and some demographic variables (e.g. SES) and relies on the assumption that
SES measured at the time of the study reflects SES in adolescence and young adulthood
when IM would have occurred. These types of information bias are likely to have
resulted in non-differential misclassification of the exposure with respect to the outcome,
biasing results toward the null. In addition, the strongly positive trends in the expected
direction (SES, sibship size) suggest a true association.

57
Although the results can be generalized to adults of European descent living in the
United States and Europe, the limited number of ethnically diverse participants enrolled
in these studies and the exclusion of HIV/AIDS-related lymphomas and post-transplant
lymphomas limits generalizability to other groups. Because NHL patients were recruited
after the onset of disease, those with longer post-diagnosis survival times were more
likely to enroll in the study and complete questionnaires. This ascertainment bias
prevents us from generalizing to NHL patients with very short survival times, although
rapid case ascertainment methods at individual study sites dampened the impact of this
bias. In general, survival times for TCL patients are shorter than those for B-cell
lymphoma patients (163,164). Among our sample of TCL participants, the majority were
diagnosed with peripheral T-cell (51%) or mycosis fungoides/Sézary syndrome (MF/SS)
(33%). Survival times for these subtypes vary significantly depending on stage at
presentation and disease-specific factors (e.g. level of skin involvement by patch or
plaque in MF/SS) (165,166). The introduction of new treatments such as Rituximab
during the recruitment window for our study may have had additional impact on the
subtypes of NHL patients we were able to recruit for study inclusion. Follow-up analyses
are warranted to determine whether the effect modification we identified applies to
patients presenting with advanced or aggressive disease.
Furthermore, data from clinic-based case-control studies are subject to Berkson’s
bias since patient controls are likely to be sicker than the general population from which
cases were ascertained. Many admitting conditions of clinic-based controls may have
some immune component which can obscure the effect of immune-related genetic

58
variants on NHL and IM. Results of sensitivity analyses excluding clinic-based case-
control studies were directionally consistent with results using the full dataset, indicating
the effect of Berkson’s bias on our study results was minimal.
Our study was underpowered to detect an interaction between uncommon variants
and IM within rare NHL subtype strata after adjusting for multiple comparisons. For
example, in order to achieve 80% power of detecting an interaction odds ratio of 0.09 for
rs1143627 at α = 0.05 after accounting for multiple comparisons, we would have needed
518 genotyped TCL patients. Thus, even with the overall large numbers of cases and
controls in the study, there was inadequate power to detect associations by subtype.
In summary, we confirmed the long-standing association between IM risk and
higher SES and lower sibship size (for those born prior to 1960) and showed a female
excess that requires confirmation and further explanation. This study was also the first to
explore possible interaction between immune response genotypes and IM history on NHL
risk (167). The results from our study may have broader implications for understanding
how certain genotypes modulate the impact of various infectious agents on NHL
etiology. The identified variants in IL1B and IL6 may influence T-cell activation, growth,
and differentiation in the presence of the massive T-cell expansion associated with IM
leading to decreased immune cell proliferation. Although we observed a possible
interaction that affected the risk of a rare NHL subtype, our study was underpowered to
overcome multiple comparisons. Confirmation will require a well-characterized, targeted
study with larger numbers.

59

TABLES
Table 3.1: Source of Participants from InterLymph Case-Control Studies: EpiLymph
NHL Patients Controls
Study Location Year
Age
(years)
Matching
variables Questionnaire
Sample
type
N (% IM
history
available)

N (% IM
history
available) Source
Source
type
EpiLymph Spain 1998–2003 17-96 Age,
sex,
region
Self-
administered
Blood 441 (99%)

631 (99%) Patients admitted to hospital for
infectious, parasitic,
mental, nervous, circulatory,
digestive, endocrine,
metabolic or respiratory conditions
clinic-
based

France 2000–2003 18–82 Age,
sex,
region
Self-
administered
Blood 218 (100%)

276 (100%) Patients admitted to hospital for
infectious, parasitic,
mental, nervous, circulatory,
digestive, endocrine,
metabolic and respiratory
conditions
clinic-
based

Germany 1999–2002 18–82 Age,
sex,
region
Self-
administered
Blood 518 (93%)

710 (95%) Random selection from population
registries
population-
based

Italy 1998–2004 25–81 Age,
sex,
region
Self-
administered
Blood 222 (99.5%)

336 (100%) Random selection from population
registries
population-
based

Ireland 1998–2004 19–85 Age,
sex,
center
Self-
administered
Blood 146 (97%)

208 (99.5%) Patients admitted to hospital for
infectious, parasitic,
mental, nervous, circulatory,
digestive, endocrine,
metabolic or respiratory conditions
clinic-
based
Czech
Republic
2001–2003 19–82 Age,
sex,
region
Self-
administered
Blood 199 (99%) 304 (100%) Patients admitted to hospital for
infectious, parasitic,
mental, nervous, circulatory,
digestive, endocrine,
metabolic and respiratory
conditions
clinic-
based

60

Table 3.2. Source of Participants from InterLymph Case-Control Studies: BC, Mayo Clinic, NCI-SEER, Scale, UCSF, Yale
NHL Patients Controls
Study Location Year
Age
(years)
Matching
variables Questionnaire Sample type
N (% IM
history
available)

N (% IM
history
available) Source
Source
type
BC British Columbia,
Canada
2000–2004 20–80 Age,
sex,
region
Self-
administered
and
computer-
assisted
telephone
interview
Blood or saliva 833 (93%)

848 (99%) Random selection
from client registry
of Ministry of
Health
population-
based
Mayo Clinic Minnesota 2002–2008 20–87 Age,
sex,
region
Self-
administered
Blood 1128 (81%)

1319 (81%) Patients attending
a pre-scheduled
general medical
exam
clinic-
based
NCI-SEER Detroit; Iowa;
Los Angeles;
Seattle, USA
1998–2001 20–74 Age,
region,
race
Self-
administered
Blood and Saliva 1321 (57%)

1057 (57%) <65y: RDD; 65y+:
random selection
from CMS
population-
based
SCALE Denmark; Sweden 1999–2002 18–74 Age,
sex,
country
Telephone
interview with
standardized,
computer-
aided
questionnaire
Blood 3055 (99%)

3187 (99%) Random selection
from population
registries
population-
based
UCSF1 San Francisco,
CA, USA
1988–1995 21–74 Age,
sex,
region
In-person
Interview
Blood 1302 (99.8%)

2402 (99.8%) All ages:
RDD;plus for 65y:
random selection
from
CMS
population-
based
Yale Connecticut, USA 1995–2001 21–84 Age, sex In-person
Interview
Blood 600 (1%) 717 (100%) <65y: RDD; 65y+:
random selection
from CMS
population-
based

61
Table 3.3. Demographic Characteristics of 7926 NHL Patients
Negative IM history Positive IM History

N (%)

N (%)
Study Center BC 566 (8%)

42 (10%)

EpiLymph-Czech Republic 165 (2%)

5 (1%)

EpiLymph-France 198 (3%)

3 (1%)

EpiLymph-Germany 435 (6%)

18 (4%)

EpiLymph-Ireland 116 (2%)

11 (3%)

EpiLymph-Italy 177 (2%)

2 (0%)

EpiLymph-Spain 418 (6%)

6 (1%)

Mayo Clinic 779 (10%)

80 (20%)

NCI-SEER 543 (7%)

48 (12%)

SCALE 2,653 (35%)

94 (23%)

UCSF 946 (13%)

63 (15%)

Yale 523 (7%)

35 (9%)

SES
a
Low (0-12 years) 3,037 (40%)

74 (18%)

Medium 2,400 (32%)

134 (33%)

High 2,082 (28%)

199 (49%)

Birth order First/Only 2,555 (34%)

149 (37%)

2nd 1,781 (24%)

115 (28%)

3rd 1,031 (14%)

50 (12%)

4th 1,392 (19%)

40 (10%)

Missing 760 (10%)

53 (13%)

Number of Siblings 0 255 (3%)

23 (6%)

1 1,144 (15%)

71 (17%)

2 1,603 (21%)

118 (29%)

3 3,908 (52%)

153 (38%)

Missing 609 (8%)

42 (10%)

Sex Male 4,052 (54%)

186 (46%)

Female 3,467 (46%)

221 (54%)

Mean ± SD Med (IQR)

Mean ± SD Med (IQR)
Age at NHL Diagnosis/Interview 60 ± 12 62 (17) 52 ± 13 53 (19)
IM: infectious mononucleosis; IQR: interquartile range; NHL: non-Hodgkin lymphoma; SD: standard deviation; SES:
socioeconomic status

a
Socioeconomic status (SES) was categorized based on years of education (low: 0-12 years, high school or less;
medium: 13-15 years, some college; high: 16+ years, college degree or more) or tertiles of the SES variable submitted
by each individual study center.

62
Table 3.4. Demographic Characteristics of 10018 Controls
Negative IM history Positive IM History

N (%)

N (%)
Study Center BC 604 (6%)

35 (7%)

EpiLymph-Czech Republic 289 (3%)

8 (2%)

EpiLymph-France 250 (3%)

5 (1%)

EpiLymph-Germany 628 (7%)

21 (4%)

EpiLymph-Ireland 198 (2%)

5 (1%)

EpiLymph-Italy 331 (3%)

3 (1%)

EpiLymph-Spain 603 (6%)

5 (1%)

Mayo Clinic 1,014 (11%)

85 (16%)

NCI-SEER 378 (4%)

25 (5%)

SCALE 2,830 (30%)

106 (20%)

UCSF 1,752 (18%)

189 (36%)

Yale 620 (7%)

34 (7%)

SES
a
Low (0–12 years) 3,282 (35%)

63 (12%)

Medium (13–15 years) 3,169 (33%)

146 (28%)

High (16+ years) 3,046 (32%)

312 (60%)

Birth order First/Only 3,318 (35%)

182 (35%)

2nd 2,412 (25%)

154 (30%)

3rd 1,278 (13%)

83 (16%)

4th 1,653 (17%)

57 (11%)

Missing 836 (9%)

45 (9%)

Number of Siblings 0 394 (4%)

24 (5%)

1 1,578 (17%)

110 (21%)

2 2,147 (23%)

159 (31%)

3 4,673 (49%)

189 (36%)

Missing 705 (7%)

39 (7%)

Sex Male 5,018 (53%)

260 (50%)

Female 4,479 (47%)

261 (50%)

Mean ± SD Med (IQR)

Mean ± SD Med (IQR)
Age at NHL Diagnosis/Interview 57 ± 15 60 (21)

46 ± 15 47 (22)
IM: infectious mononucleosis; IQR: interquartile range; NHL: non-Hodgkin lymphoma; SD: standard deviation; SES:
socioeconomic status
a
Socioeconomic status (SES) was categorized based on years of education (low: 0-12 years, high school or less;
medium: 13–15 years, some college; high: 16+ years, college degree or more) or tertiles of the SES variable submitted
by each individual study center.

63
Table 3.5. Subtypes Among NHL Patients
N (%)
B-cell DLBCL 2246 (28%)

CLL/SLL/PLL/MCL 1470 (19%)

Follicular 1691 (21%)

MZL 447 (6%)

MCL 325 (4%)

LPL/Waldenstrom 228 (3%)

Hairy cell 75 (1%)

Burkitt 63 (1%)

Precursor B-cell 40 (1%)

Burkitt-like 27 (0.3%)

B-Cell NOS 534 (7%)

TOTAL B-Cell 7146 (90%)

T-Cell Peripheral T-cell 262 (3%)

MF/SS 166 (2%)

Precursor T-cell 26 (0.3%)

Nasal NK 17 (0.2%)

Large granular 7 (0.1%)

T-PLL 4 (0.1%)

T-Cell NOS 27 (0.3%)

TOTAL T-Cell 509 (6%)

NOS

210 (3%)
Missing
a
61 (1%)
B-PLL: B-cell prolymphocytic leukemia; CLL/SLL: chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL:
diffuse large B-cell lymphoma; FL: follicular lymphoma; LPL: lymphoplasmacytic lymphoma; MCL: mantle cell
lymphoma; MF/SS: mycosis fungoides/Sézary syndrome; MZL: marginal zone lymphoma; NK: natural killer; NHL: non-
Hodgkin lymphoma; NOS: not otherwise specified; T-PLL: T-cell prolymphocytic leukemia.
a
Patients missing a subtype were excluded from subtype-specific analyses

64
Table 3.6. Associations Between NHL and Candidate Risk Variants [IL1A (rs1800587),
IL1B (rs16944, rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6 (rs1800795,
rs1800797), IL10 (rs1800896, rs1800890), TNFA (rs1800629), HLA I (rs6457327), and
HLA II (rs10484561)]

Genotyped
Controls (N)
Genotyped
NHL Patients (N) OR
a
95% CI P value
IL1A-889C>T (rs1800587) 2,317 2,084 1.00 0.89, 1.13 0.97

IL1B-511C>T (rs16944) 1,280 1,311 1.17 1.00, 1.37 0.05

IL1B-31C>T (rs1143627) 3,715 3,130 1.06 0.96, 1.17 0.22

IL1RN-9589A>T (rs454078) 2,319 2,068 1.19 1.06, 1.35 0.004 **
IL2–384T>G (rs2069762) 2,320 2,080 1.11 0.99, 1.25 0.09

IL6-174G>C (rs1800795) 2,347 2,099 0.93 0.82, 1.05 0.25

IL6-597G>A (rs1800797) 3,852 3,304 0.94 0.85, 1.04 0.22

IL10-1082A>G (rs1800896) 4,173 3,472 1.06 0.96, 1.18 0.24

IL10-3575T>A (rs1800890) 5,914 5,629 1.08 1.00, 1.16 0.06

TNF-308G>A (rs1800629) 5,562 5,546 1.12 1.03, 1.22 0.01 *
HLA: C>A (rs6457327) 2,963 2,457 0.82 0.74, 0.92 <0.001 **
HLA: T>G (rs10484561) 3,989 3,176 1.31 1.17, 1.46 <0.001 **
CI: confidence interval; NHL: non-Hodgkin lymphoma; OR: odds ratio.
* P value < 0.05 before adjusting for multiple comparisons, ** PACT < 0.05 after adjustment for multiple comparisons.
a
ORs, CIs, and P values calculated using logistic regression models adjusted for age, sex, study center, and
socioeconomic status

65
Table 3.7. Association Between NHL and Infectious Mononucleosis by NHL Subtype

Controls (N) NHL Patients (N) OR
a
95% CI P value

All NHL 10,018 7,926 1.20 1.04, 1.38 0.01 *
CLL/SLL 10,018 1,466 1.51 1.13, 2.02 0.006 **
DLBCL 10,018 2,246 1.10 0.89, 1.37 0.37

FL 10,018 1,691 1.08 0.86, 1.37 0.49

MCL 10,018 325 2.29 1.41, 3.75 0.001 **
TCL 10,018 509 1.32 0.89, 1.96 0.17
CI: confidence interval; CLL/SLL: chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL: diffuse large B-
cell lymphoma; FL: follicular lymphoma; MCL: mantle cell lymphoma; NHL: non-Hodgkin lymphoma; OR: odds ratio;
TCL: T-cell lymphoma.
* P value < 0.05 before adjusting for multiple comparisons, ** P value < 0.05 after Bonferroni adjustment for multiple
comparisons.
a
ORs, CIs, and P values calculated using logistic regression models stratified by genotype and adjusted for age, sex,
study center, and socioeconomic status.

66
Table 3.8. Interaction Between Infectious Mononucleosis History and Candidate Risk
Variants [IL1A (rs1800587), IL1B (rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6
(rs1800795, rs1800797), IL10 (rs1800890), TNFA (rs1800629), HLA I (rs6457327), and
HLA II (rs10484561)] on NHL, CLL/SLL, and DLBCL risk: Empirical-Bayes Estimates
of Interaction Effects
NHL
subtype Variant
Genotyped
Controls (N)
Genotyped
NHL Patients (N)
Interaction effect
OR
a
95% CI P value
All NHL IL1A-889C>T (rs1800587) 2,317 2,084 1.13 0.69, 1.87 0.62

IL1B-511C>T (rs16944) 1,280 1,311 0.66 0.35, 1.22 0.18

IL1B-31C>T (rs1143627) 3,715 3,130 0.76 0.57, 1.02 0.06

IL1RN-9589A>T (rs454078) 2,319 2,068 1.02 0.62, 1.67 0.94

IL2–384T>G (rs2069762) 2,320 2,080 0.99 0.62, 1.57 0.96

IL6-174G>C (rs1800795) 2,347 2,099 1.08 0.77, 1.52 0.64

IL6-597G>A (rs1800797) 3,852 3,304 1.10 0.81, 1.49 0.55

IL10-1082A>G (rs1800896) 4,173 3,472 0.81 0.59, 1.10 0.18

IL10-3575T>A (rs1800890) 5,914 5,629 0.80 0.57, 1.13 0.21

TNF-308G>A (rs1800629) 5,562 5,546 0.86 0.59, 1.27 0.46

HLA: C>A (rs6457327) 2,963 2,457 1.07 0.65, 1.76 0.78

HLA: T>G (rs10484561) 3,989 3,176 0.93 0.62, 1.40 0.74

CLL/SLL IL1A-889C>T (rs1800587) 2,317 366 1.03 0.45, 2.35 0.95

IL1B-511C>T (rs16944) 1,280 117 1.30 0.22, 7.56 0.77

IL1B-31C>T (rs1143627) 3,715 646 0.96 0.49, 1.87 0.90

IL1RN-9589A>T (rs454078) 2,319 364 1.81 0.80, 4.07 0.15

IL2–384T>G (rs2069762) 2,320 365 1.77 0.79, 3.98 0.17

IL6-174G>C (rs1800795) 2,347 364 1.13 0.49, 2.57 0.78

IL6-597G>A (rs1800797) 3,852 666 0.89 0.45, 1.76 0.73

IL10-1082A>G (rs1800896) 4,173 669 0.80 0.39, 1.62 0.53

IL10-3575T>A (rs1800890) 5,914 1,204 0.68 0.37, 1.24 0.21

TNF-308G>A (rs1800629) 5,562 1,186 0.89 0.47, 1.71 0.73

HLA: C>A (rs6457327) 2,963 389 1.05 0.36, 3.02 0.93

HLA: T>G (rs10484561) 3,989 623 0.54 0.20, 1.50 0.24

DLBCL IL1A-889C>T (rs1800587) 2,317 541 0.98 0.48, 2.01 0.96

IL1B-511C>T (rs16944) 1,280 384 0.75 0.34, 1.68 0.49

IL1B-31C>T (rs1143627) 3,715 877 0.61 0.34, 1.08 0.09

IL1RN-9589A>T (rs454078) 2,319 530 0.83 0.41, 1.68 0.61

IL2–384T>G (rs2069762) 2,320 538 2.02 0.99, 4.13 0.05

IL6-174G>C (rs1800795) 2,347 537 0.83 0.43, 1.60 0.59

IL6-597G>A (rs1800797) 3,852 922 0.92 0.52, 1.64 0.78

IL10-1082A>G (rs1800896) 4,173 928 1.10 0.58, 2.09 0.78

IL10-3575T>A (rs1800890) 5,914 1,496 0.74 0.45, 1.21 0.23

TNF-308G>A (rs1800629) 5,562 1,447 0.72 0.42, 1.23 0.23

HLA: C>A (rs6457327) 2,963 701 0.66 0.32, 1.37 0.26

HLA: T>G (rs10484561) 3,989 840 1.05 0.51, 2.17 0.89
CI: confidence interval; CLL/SLL: chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL: diffuse large B-
cell lymphoma; NHL: non-Hodgkin lymphoma; OR: odds ratio.
a
Interaction ORs, CIs, and P values calculated using empirical-Bayes method adjusted for age, sex, study center, and
socioeconomic status.

67
Table 3.9. Interaction Between Infectious Mononucleosis History and Candidate Risk
Variants [IL1A (rs1800587), IL1B (rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6
(rs1800795, rs1800797), IL10 (rs1800890), TNFA (rs1800629), HLA I (rs6457327), and
HLA II (rs10484561)] on FL, MCL, and TCL risk: Empirical-Bayes Estimates of
Interaction Effects

Genotyped
Controls (N)
Genotyped
NHL Patients (N)
Interaction Effect
NHL
subtype Variant OR
a
95% CI P value
FL IL1A-889C>T (rs1800587) 2,317 527 1.18 0.58, 2.41 0.64

IL1B-511C>T (rs16944) 1,280 331 0.53 0.20, 1.36 0.19

IL1B-31C>T (rs1143627) 3,715 706 0.98 0.54, 1.77 0.95

IL1RN-9589A>T (rs454078) 2,319 526 0.63 0.31, 1.29 0.21

IL2–384T>G (rs2069762) 2,320 528 0.69 0.35, 1.35 0.28

IL6-174G>C (rs1800795) 2,347 533 1.34 0.69, 2.60 0.39

IL6-597G>A (rs1800797) 3,852 757 1.78 0.94, 3.39 0.08

IL10-1082A>G (rs1800896) 4,173 750 0.67 0.37, 1.20 0.18

IL10-3575T>A (rs1800890) 5,914 1,125 0.89 0.51, 1.54 0.68

TNF-308G>A (rs1800629) 5,562 1,130 0.91 0.50, 1.65 0.75

HLA: C>A (rs6457327) 2,963 510 1.65 0.72, 3.79 0.24

HLA: T>G (rs10484561) 3,989 696 0.86 0.46, 1.62 0.64

MCL IL1A-889C>T (rs1800587) 2,317 103 2.24 0.48, 10.33 0.30

IL1B-511C>T (rs16944) 1,280 61 0.25 0.02, 3.07 0.28

IL1B-31C>T (rs1143627) 3,715 146 1.35 0.33, 5.57 0.68

IL1RN-9589A>T (rs454078) 2,319 102 1.24 0.28, 5.53 0.78

IL2–384T>G (rs2069762) 2,320 103 0.74 0.16, 3.43 0.70

IL6-174G>C (rs1800795) 2,347 105 0.28 0.06, 1.24 0.09

IL6-597G>A (rs1800797) 3,852 159 0.29 0.08, 1.11 0.07

IL10-1082A>G (rs1800896) 4,173 171 0.70 0.15, 3.22 0.64

IL10-3575T>A (rs1800890) 5,914 285 0.59 0.20, 1.77 0.35

TNF-308G>A (rs1800629) 5,562 279 3.27 1.06, 10.05 0.04

HLA: C>A (rs6457327) 2,963 116 2.75 0.34, 22.05 0.34

HLA: T>G (rs10484561) 3,989 158 0.29 0.04, 2.33 0.24

TCL IL1A-889C>T (rs1800587) 2,317 127 2.76 0.36, 21.42 0.33

IL1B-511C>T (rs16944) 1,280 97 0.01 0.00, 2.55 0.10

IL1B-31C>T (rs1143627) 3,715 206 0.09 0.01, 0.87 0.04 *

IL1RN-9589A>T (rs454078) 2,319 125 0.37 0.05, 2.75 0.33

IL2–384T>G (rs2069762) 2,320 127 0.71 0.12, 4.36 0.71

IL6-174G>C (rs1800795) 2,347 129 0.05 0.00, 1.01 0.05

IL6-597G>A (rs1800797) 3,852 219 0.08 0.01, 0.80 0.03 *

IL10-1082A>G (rs1800896) 4,173 233 0.89 0.16, 5.04 0.90

IL10-3575T>A (rs1800890) 5,914 378 1.07 0.36, 3.16 0.91

TNF-308G>A (rs1800629) 5,562 369 1.20 0.39, 3.71 0.75

HLA: C>A (rs6457327) 2,963 183 3.36 0.48, 23.47 0.22

HLA: T>G (rs10484561) 3,989 210 1.82 0.31, 10.58 0.50

CI: confidence interval; FL: follicular lymphoma; MCL: mantle cell lymphoma; NHL: non-Hodgkin lymphoma; OR: odds
ratio; TCL: T-cell lymphoma.
* P value < 0.05 before adjusting for multiple comparisons. Significance not retained after accounting for multiple
comparisons using PACT statistic.
a
Interaction ORs, CIs, and P values calculated using empirical-Bayes method adjusted for age, sex, study center, and
socioeconomic status.

68
Table 3.10. Association Between Infectious Mononucleosis History and T-Cell
Lymphoma Among Genotyped Participants Stratified by IL1B (rs16944, rs1143627) and
IL6 (rs1800795, rs1800797) Genotypes
Variant
Controls

T-Cell Patients
Genotype IM- IM+ IM- IM+ OR
a
95% CI P value
IL1B-511C>T (rs16944) CC 559 35

42 5 1.91 0.67, 5.45 0.23

TC/TT 654 32

47 3 1.39 0.38, 5.11 0.62

IL1B-31C>T (rs1143627) TT 1,626 86

95 7 1.55 0.67, 3.62 0.31

CT/CC 1,907 96

100 4 0.92 0.32, 2.63 0.87

IL6-174G>C (rs1800795) GG 734 54

46 5 1.25 0.45, 3.50 0.67

CG/CC 1,461 98

74 4 0.78 0.27, 2.26 0.64

IL6-597G>A (rs1800797) GG 1298 66

73 6 1.30 0.51, 3.27 0.58
AG/AA 2,364 124 135 5 0.77 0.30, 1.98 0.59
CI: confidence interval; OR: odds ratio.
a
ORs, CIs, and P values calculated using logistic regression models stratified by genotype and adjusted for age, sex,
study center, and socioeconomic status. Effect estimates are unstable due to low sample sizes in IM+ T-cell patients’
strata.

69
Table 3.11. Association Between Age at NHL Diagnosis and Infectious Mononucleosis
Not adjusted for birth year

Adjusted for birth year

N β
a
95% CI

β
b
95% CI P value
All NHL 7926 -6.59 -7.80, -5.37

-0.07 -0.22, 0.07 0.30
CLL/SLL 1466 -6.04 -8.42, -3.66

-0.31 -0.67, 0.05 0.09
DLBCL 2246 -5.7 -8.20, -3.20

-0.12 -0.38, 0.14 0.35
FL 1691 -6.92 -9.25, -4.58

0.25 -0.05, 0.56 0.10
MCL 325 -6.66 -11.10, -2.23

-0.58 -1.18, 0.02 0.06
TCL 509 -12.11 -17.35, -6.86 -0.37 -0.92, 0.19 0.19
CI: confidence interval; CLL/SLL: chronic lymphocytic leukemia/small lymphocytic lymphoma; DLBCL: diffuse large B-
cell lymphoma; FL: follicular lymphoma; MCL: mantle cell lymphoma; NHL: non-Hodgkin lymphoma; TCL: T-cell
lymphoma.
a
Adjusted for age, sex, study center, socioeconomic status
b
Adjusted for age, sex, study center, socioeconomic status, year of birth

70
Table 3.12. Associations Between Infectious Mononucleosis History and Demographic
Factors Among 10018 Controls
IM negative
controls (N)
IM positive
controls (N)

OR
b
95% CI P value

SES Low SES 3,282 63 (ref)

Med SES 3,169 146 1.69 1.24, 2.30 0.001 **

High SES 3,046 312 2.86 2.13, 3.82 <0.001 **

Trend 9,497 521 1.69 1.48, 1.93 <0.001 **

Sex Male 5,018 260 (ref)

Female 4,479 261 1.53 1.26, 1.87 <0.001 **

CI: confidence interval; IM: infectious mononucleosis; OR: odds ratio; ref: reference category; SES: socioeconomic
status.

** P value < 0.05 after Bonferroni adjustment for multiple comparisons.

a
Socioeconomic status (SES) was categorized based on years of education (low: 0-12 years, high school or less;
medium: 13-15 years, some college; high: 16+ years, college degree or more) or tertiles of the SES variable
submitted by each individual study center.

b
ORs, CIs, and P values calculated using logistic regression models adjusted for age and study center. Models for
SES, number of siblings, and birth order were further adjusted for sex. Model for sex was further adjusted for SES.

71
Table 3.13. Association Between Infectious Mononucleosis History and Family Structure
Among Controls with Available Year of Birth Data
Year of birth
a

IM negative
controls (N)
IM positive
controls (N) OR
d
95% CI P value
All Number of
Siblings
b

0 or 1 933 54 (ref) --

2 899 56 1.03 0.69, 1.55 0.88

3+ 2,067 56 0.57 0.38, 0.85 0.01 **

Birth Order
c
1st 750 30 (ref) --

(among controls
with 2+ siblings)
2nd 745 36 1.41 0.85, 2.36 0.19

3rd+ 1,447 46 0.98 0.60, 1.59 0.92

1901–1960 Number of
Siblings
b

0 or 1 733 34 (ref) --

2 685 34 1.10 0.66, 1.82 0.71

3+ 1,822 35 0.40 0.24, 0.67 <0.001 **

Birth Order
c
1st 602 21 (ref)

(among controls
with 2+ siblings)
2nd 640 22 1.04 0.56, 1.94 0.89

3rd+ 1,244 26 0.61 0.34, 1.12 0.11

1961–1990 Number of
Siblings
b

0 or 1 200 20 (ref)

2 213 22 1.01 0.52, 1.97 0.99

3+ 245 21 1.07 0.53, 2.17 0.85

Birth Order
c
1st 147 9 (ref)

(among controls
with 2+ siblings)
2nd 105 14 --
e

3rd+ 203 20 --
e

CI: confidence interval; IM: infectious mononucleosis; OR: odds ratio; ref: reference category.
** P value < 0.05 after Bonferroni adjustment for multiple comparisons.
a
Year of birth was for controls was reported at eight study sites (38% of controls)
b
Sibship size was not reported for 0.4% of controls at sites with year of birth data
c
Birth order was not reported for 1% of controls at sites with year of birth data
d
ORs, CIs, and P values calculated using logistic regression models adjusted for age, study center, and sex.
e
Insufficient data to compute effect estimate.

72
Table 3.14. Associations Between Infectious Mononucleosis and Candidate Risk Variants
[IL1A (rs1800587), IL1B (rs1143627), IL1RN (rs454078), IL2 (rs2069762), IL6
(rs1800797), IL10 (rs1800890), TNFA (rs1800629), HLA I (rs6457327), and HLA II
(rs10484561)] Among Controls with Available Genetic Data

IM negative
controls (N)
IM positive
controls (N) OR
a
95% CI P value
IL1A-889C>T (rs1800587) 2,168 149 0.72 0.51, 1.02 0.07
IL1B-511C>T (rs16944) 1,213 67 0.76 0.46, 1.27 0.30
IL1B-31C>T (rs1143627) 3,533 182 1.03 0.75, 1.40 0.86
IL1RN-9589A>T (rs454078) 2,169 150 0.76 0.53, 1.08 0.13
IL2–384T>G (rs2069762) 2,173 147 1.25 0.88, 1.76 0.22
IL6-174G>C (rs1800795) 2,195 152 0.94 0.66, 1.34 0.73
IL6-597G>A (rs1800797) 3,662 190 0.95 0.69, 1.30 0.73
IL10-1082A>G (rs1800896) 3,975 198 1.04 0.75, 1.46 0.81
IL10-3575T>A (rs1800890) 5,663 251 1.16 0.88, 1.53 0.29
TNF-308G>A (rs1800629) 5,318 244 1.22 0.92, 1.62 0.16
HLA: C>A (rs6457327) 2,858 105 1.22 0.80, 1.86 0.36
HLA: T>G (rs10484561) 3,801 188 0.88 0.59, 1.31 0.52
CI: confidence interval; IM: infectious mononucleosis; OR: odds ratio.
a
Adjusted for age, sex, study center, socioeconomic status

73
Chapter 4. A GENOME-WIDE ASSOCIATION STUDY OF EPSTEIN-BARR
VIRUS VIRAL COPY NUMBER AND ANTIBODIES TO EPSTEIN-BARR VIRUS
ANTIGENS AMONG HODGKIN LYMPHOMA SURVIVORS
ABSTRACT
We explored the effect of genetic variation on Epstein-Barr virus (EBV) copy
number and antibody levels to EBV antigens among a set of European origin Hodgkin
lymphoma (HL) survivors from two studies. The University of Southern California
(USC) and the Université de Paris René Descartes (UPRD) contributed 166 and 106
participants, respectively. Existing GWAS data from each study were imputed separately
using the Michigan Imputation Server (Minimac4). To assess the association between the
genetic variants and EBV viral copy number, log-transformed linear regression adjusted
for sex and the first ten principal components was performed separately and then
combined using a fixed effects meta-analysis. The same linear regression method was
repeated in the USC subset with antibody data (N = 143). We identified a missense
variant at 3q29 in MUC4 associated with EBV viral load (rs2246901; 95% confidence
interval (CI): 473, 1572; P = 7.22 × 10
-9
). An intronic variant at 3p26 in GRM7
(rs140444865; 95% CI: 0.20, 0.43; P = 5.47 × 10
-9
) and an intergenic variant between
INHA and STK11IP at 2q35 (rs115805790; 95% CI: 0.17, 0.41; P = 2.70 × 10
-8
) were
associated with antibody levels to EBV nuclear antigen 1.

74
Epstein-Barr virus (EBV) (2) is a ubiquitous human herpes virus which is
estimated to infect over 90% of adults worldwide. EBV is most commonly spread via
saliva, and primary infection typically occurs sub-clinically during early childhood. In
populations of high socioeconomic status (SES), primary infection frequently occurs in
adolescence or young adulthood and may result in infectious mononucleosis (IM), a
clinical syndrome of varying severity characterized by fever, tonsillar pharyngitis, and
lymphadenopathy (3). Following primary infection, the virus persists throughout the life
of the host by immortalizing B cells and establishing transcriptionally quiescent latency
in which only non-coding RNAs are expressed. Occasionally, EBV reactivates and
reenters the lytic phase of infection during which the virus sheds and can be transmitted
to susceptible individuals. This cycle between lytic and latent infection typically persists
benignly, but EBV has been associated with several malignancies.
A substantial body of evidence suggests EBV plays a role in the pathogenesis of
classic Hodgkin lymphoma (HL), a cancer characterized by clonal tumor cells including
large, multinucleated cells and mononuclear cells —Hodgkin Reed-Sternberg cells (HRS)
—in an inflammatory background (55). Compared to patients with other lymphomas, HL
patients have elevated EBV antibody titers (57) several years before HL diagnosis (58).
Viral DNA and EBV proteins are present in HRS cells of ~40% of HL cases in the
United States and Europe (59,60). In a cohort study of HL, risk of EBV-positive HL was
elevated in participants who had been diagnosed with IM while risk of EBV-negative HL
was not elevated after IM diagnosis (61). Frequency of EBV-positivity varies
significantly by HL histologic subtype (nodular sclerosis: 20–40%; mixed cellularity: 60–

75
75%; lymphocyte-depleted: 80–90%, lymphocyte rich: <10%) and patient characteristics
(Hispanic ethnicity, male sex, low SES, immunosuppression, very young or very old age
at diagnosis) (6,62).
Previous genome-wide association studies (GWAS), including the two studies
contributing data to this paper, have identified several loci significantly associated with
HL risk and prognosis. Some loci are associated with all HL, and others are associated
with tumor EBV status or histological subtype (63,168–171). A previous study
comparing EBV viral copy number in 32 pairs of twins discordant for HL showed that
the HL survivor (case twin) had higher EBV copy number levels than the unaffected
twin. This elevated viral load was observed among all histopathological subtypes,
including nodular sclerosis, which is not usually associated with EBV+ tumors
(172). Elevated anti-viral capsid antigen (VCA) immunoglobulin (Ig) G titers observed
among healthy first-degree relatives (but not spouses) of HL patients suggested a possible
role for genetics in EBV control among HL patients (173). Genetic variation in the HLA
region is associated with anti-EBV nuclear antigen 1 (EBNA-1) IgG levels (174). Among
HL patients and their first degree relatives, correlation between EBV viral load in
peripheral blood mononuclear cells and anti-VCA IgG levels were also observed (175).
EBV may also influence prognosis among HL patients. Associations between higher
pretreatment plasma EBV viral genome copy number (derived from the tumor cells in
EBV-positive cases) and lower failure-free HL survival rates have been reported (67).
Here, we investigate whether there are genetic determinants of control and
response to EBV infection among HL survivors. Our GWAS and meta-analyses combine

76
existing genetic and EBV viral copy number data with newly generated EBV serology
data.
MATERIALS AND METHODS
This study was approved by the institutional review board of the Keck School of
Medicine of USC and the appropriate French Consultative Committee Protecting Persons
in Biomedical Research in accordance with the Declaration of Helsinki. Signed informed
consent was obtained from all adult participants and the parents of all minor participants.
Participant recruitment, genotyping, and EBV quantification methods have been
described previously (170,175,176). A summary is provided below.
Participants
Data contributed by two study sites were combined for this study: one from the
University of Southern California (USC) and the other from Université de Paris René
Descartes (UPRD).
USC: Incident HIV-negative HL survivors were recruited from the USC Cancer
Surveillance Program and the Cancer Prevention Institute of California (the Los Angeles
County and Greater San Francisco Bay Area Survey of Epidemiology and End Results
registries, respectively) (170,177–179) as well as the population-based California Twin
Program (94) and volunteer International Twin Study (180). Participants were recruited
for study participation between 2006 and 2010 and were 14–52 years of age (median =
28.3) at the time of HL diagnosis. Race/ethnicity, date of birth, and IM history were
collected via questionnaire. HL histology, tumor EBV status, and date of diagnosis were

77
abstracted from medical records. Cancer registry participants’ samples and information
were collected 1–9 years after diagnosis (median = 3.6) and twin participants’ after 8–42
years (median = 24). Additional details about the USC discovery set of HL survivors can
be found in previous publications (64,170,172).
UPRD: HIV-negative HL survivors in complete remission were recruited from
three hematology units in the Paris region (Saint-Louis, Necker, and Gustave Roussy
hospitals) at least 1 year after diagnosis. Participants were recruited for study
participation between 2002 and 2005 and were between 8 and 47 years of age (median =
25) at the time of HL diagnosis. HL histology, tumor EBV status, IM history, and date of
diagnosis were abstracted from medical records, specific interviews, and pathological
review. French legal restrictions prevented UPRD investigators from collecting self-
reported ethnicity data. Additional details about the UPRD discovery set of HL survivors
can be found in previous publications (174,175).
When data were available for multiple HL survivors from the same family, one
family member was randomly selected for inclusion. Participants were excluded from the
study if genotyping data were unavailable. Participants who did not contribute blood
samples (saliva only) were also excluded from this study. Participants with insufficient
plasma available for antibody testing were excluded from analyses of corresponding
phenotypes.
Genotyping and Imputation
USC participants’ DNA was isolated from whole blood using QIAamp 96 DNA
Blood Mini Kits (USC Genomics Core/QIAGEN, Hilden, Germany) or from saliva using

78
Oragene saliva self-collection kits (DNA Genotek, Ottawa, Canada), and participants
were genotyped using the Illumina Human610-Quad Beadchip in accordance with
manufacturer’s guidelines (Illumina, California, USA). UPRD participants’ genomic
DNA was isolated from blood samples, using QIamp DNA Blood Mini Kits (QIAGEN,
Hilden, Germany), and participants were genotyped using the Illumina HumanCy-
toSNP12v2.1 Panel. Prior to imputation, genotype data from each study site were cleaned
separately to remove variants according to the following exclusion criteria:
monomorphic, minor allele frequency (MAF) < 0.01, call rate < 95%, or strong deviation
from Hardy-Weinberg equilibrium (P < 1 ×10
-5
). We also removed all participants with a
genotyping call rate < 95%.
Following quality checks, variant formats (strand, id names, positions, alleles,
ref/alt assignment) were updated to match the reference panel (The Haplotype Reference
Consortium CEU population release 1.1, www.haplotype-reference-consortium.org).
USC and UPRD datasets were imputed separately on the Michigan Imputation Server
(https://imputationserver.sph.umich.edu/) running Minimac4 (181). After imputation,
variants with imputation quality < 0.3, minor allele frequency (MAF) < 0.05 among
participants with available phenotype data, or strong deviation from Hardy-Weinberg
equilibrium (P < 1 ×10
-6
) were removed.
EBV viral copy number. For USC participants, DNA was extracted from frozen
peripheral blood mononuclear cells (PBMCs) and shipped to the Frederick National
Laboratory for Cancer Research for EBV DNA viral load measurement, which was
determined using a real-time quantitative polymerase chain reaction (PCR) assay,

79
targeting the EBV polymerase gene. Endogenous retrovirus-3 (ERV-3) was used as a cell
quantitation marker to estimate the total number of viral copies per million cells (182).
Triplicate reactions using approximately 250 μl total DNA per reaction were performed
for EBV and ERV-3, and the average was used to calculate EBV copies per million cell
equivalents. A qualitative positive equivalent to three copies was assigned to any sample
in which only one of the three wells amplified. For low positive samples in which at least
2 wells of 3 amplified, viral load was estimated using ABI software.
For UPRD participants, PBMCs were isolated from blood samples on Ficoll-
Paque (Pharmacia Bio-tech), and pellets of 10
6
cells were frozen at -80°C. DNA was
extracted from cell pellets by use of a QIAmp blood kit (QIAGEN Inc., Courtaboeuf,
France). Viral load in PBMCs was determined by real-time quantitative PCR with a
fluorogenic probe (175). For preparation of the standard curve, the 121-bp PCR product
was directly inserted into a pcDNA 3.1 vector (Invitrogen, Groningen, The Netherlands)
containing one copy of the EBV PCR target region. Tenfold serial dilutions in water (10
6

to 10 copies) were prepared. The same preparations were used in each test. Real-time
quantitative PCR for EBV and albumin DNA quantification were carried out
simultaneously to determine the amount of cellular DNA in each sample. Results from
both sites were reported per million cells.
EBV antibody serology. Serology for EBV antigens was performed in the
Waterboer Laboratory, German Cancer Research Center, Heidelberg, Germany. Among
USC participants with a sufficient volume of blood plasma available for testing, the
presence and quantity of EBV-specific antibodies in serum were assessed using HHV

80
species-specific (‘Monoplex’) assays described by Brenner et al. (183). A panel of four
EBV proteins [EBNA-1, VCA p18, Z-Epstein-Barr virus replication activator (ZEBRA),
and early antigen-diffuse (EA-D)] was used to evaluate EBV positivity. Using protein-
specific median fluorescence intensity (MFI) thresholds (EBNA-1: 411, VCA p18: 2526,
ZEBRA: 74, EA-D: 110), antibody levels were dichotomized as positive (met or
exceeded corresponding MFI threshold) or negative (less than corresponding MFI
threshold). Participants were then classified as EBV positive if the MFI of two or more
EBV antibodies met or exceeded corresponding MFI thresholds.
Data Analysis
Kruskal-Wallis and Wilcoxon rank sum tests were used to compare EBV
phenotypes (viral load and antibody levels) across participant characteristics. Genome-
wide association analysis was performed using linear regression (EBV viral load,
antibody levels among participants exceeding MFI threshold) and logistic regression
(antibody positivity) assuming an additive model of inheritance. All models were
adjusted for sex and the first ten principal components of the variance-standardized
relationship matrix to control for shared ancestry and cryptic relatedness. EBV viral load
and EBV antibody MFI measurements were log transformed for linear regression models.
Resulting effect estimates and 95% confidence intervals (95% CI) for linear regression
models were exponentiated, and the presented estimate should be interpreted as the
proportional change in the expected geometric mean of the continuous phenotype for
each minor allele. An association was considered significant if the P value was < 5 × 10
-8
.
Results for EBV viral load (USC and UPRD) were combined via inverse variance-

81
weighted fixed-effects meta-analysis. Higgins and Thompson’s I
2
and Cochran’s Q-
statistics were calculated to determine the percentage of variation attributable to
heterogeneity and to test for heterogeneity between discovery sets in the meta-analysis.
Manhattan plots were created for all GWAS analyses, and regional plots of
statistically significant results were generated using LocusZoom (184). Quantile-quantile
(QQ) plots were produced for each GWAS analysis to examine the distribution of
observed test statistics compared to test statistics expected under the null hypothesis.
Corresponding inflation factors (l) were calculated as a ratio of the median of the
observed chi-square statistics for association from the Wald tests over the median (=
0.455) of the chi-square distribution with 1 degree of freedom (185). Sensitivity analysis
was conducted to test for effect modification by gender, IM history, HL subtype, tumor
EBV status, age at diagnosis, age at blood draw, and time between diagnosis and blood
draw in discovery sets. Annotation for mutations surpassing the threshold for statistical
significance was conducted using ANNOVAR (Annotate Variation) (186). Within
ANNOVAR, the Homo Sapiens hg19 genome assembly (hg19/GRCh37) was selected as
the reference genome, and the Reference Sequence (RefSeq) database was used for gene
definition (187,188). Statistical analysis was conducted using PLINK 1.9 (Boston,
Massachusetts, USA) (189) and R (Vienna, Austria) (95) using the qqman (190,191) and
gap (192,193) packages.

82
RESULTS
166 USC participants and 106 UPRD participants were included in the EBV viral
load meta-analysis. 143 USC participants were included in the analysis of antibodies to
EBV antigens. Characteristics of participants from USC and UPRD are described in
Table 4.1. The majority of participants were male (USC: 59%, UPRD: 59%) and of
European ethnicity (USC: 65%, UPRD: 100%). Nodular sclerosing Hodgkin lymphoma
(NSHL) was the most common subtype among our study population (USC: 79%, UPRD:
71%). Positive IM history (USC: 20%, UPRD: 19%) and positive tumor EBV status
(USC: 9%, UPRD: 19%) were reported in a minority of participants. Participants
provided blood samples an average of 6.2 years after diagnosis (median = 4.3 years).
The availability of pre-imputation and post-imputation data is described in Table
4.2. The meta-analysis included 6.8 million genotyped and imputed variants, and 7.8
million variants were included in the USC analyses of antibodies.
EBV viral load
EBV DNA was detected in a minority of HL survivors in each discovery set
(USC: 42%, UPRD: 44%). The median viral copy number was zero at both USC and
UPRD. A broader range of viral copy number values were observed at USC (P = 0.03,
Supplemental Figure 4.1). Among those with available data, there was an association
between EBV+ tumors and higher viral copy number (P = 0.03). We did not observe
differences in viral copy number by sex, histological subtype (NS or other), IM history,
age at diagnosis, age at blood draw, or time between diagnosis and blood draw (P ≥ 0.19).

83
Through meta-analysis of EBV viral load, we observed a strong signal in the
MUC4 gene at 3q29. Each A allele of rs2246901, a missense mutation genotyped at USC
and imputed at UPRD, was associated with a 1000-fold increase in the expected
geometric mean of viral copy number (P = 7.22 × 10
-9
. 95% CI: 473, 1572.). Fourteen
other variants in high LD with rs2246901 also surpassed the genome-wide significance
threshold in the meta-analysis (Table 4.3–Table 4.5, Figure 4.1, Figure 4.2). The
associations between rs2246901 and viral copy number within each discovery set did not
surpass the threshold for genome-wide significance (PUSC = 1.06 × 10
-6
, PUPRD = 5.97 ×
10
-3
).
I
2
and Q statistics suggested an absence of heterogeneity between discovery sets
(I
2
= 0%, Phomogeneity ³ 0.70). Quantile-quantile plots of discovery (Supplemental Figure
4.2) and meta-analysis results (Supplemental Figure 4.3) did not reveal overdispersion of
P values, nor did we observe inflation of the test statistic (lmeta-analysis = 1.03).
ANNOVAR was used to provide functional annotation of rs2246901 (Table 4.6).
Mutation significance predictions yielded mixed results. Mutation Assessor, PolyPhen-2
HDIV and PolyPhen-2 HVAR indicated the mutation may be functional and ‘probably
damaging’. The other significance predictions indicated the mutation is likely tolerated or
neutral. Sensitivity analyses of the association between rs2246901 and EBV viral load by
participant characteristics are described in Table 4.7 (USC) and Table 4.8 (UPRD). We
did not observe effect modification by sex, HL histology (nodular sclerosing vs. other),
tumor EBV status, self-reported IM history, age at diagnosis, age at blood draw, or time
between diagnosis and blood draw (Pinteraction > 0.20).

84
Serum Antibodies to EBV Antigens
Characteristics of the 143 USC participants with serology data available
(Supplemental Table 4.1) were similar to those of the full USC population of 166
participants. Distributions of MFI levels for EA-D, EBNA-1, VCA p18, and ZEBRA
shown in Supplemental Figure 4.4–Supplemental Figure 4.7. The median MFI for EA-D
was higher among participants with a positive IM history (P = 0.03) and those diagnosed
with nodular sclerosing HL (P = 0.04). The distribution of ZEBRA MFI varied by age at
blood draw (P = 0.04).
No associations between genetic variants and the presence of antibodies to EBV
antigens (dichotomous outcomes) met the genome-wide significance threshold
(Supplemental Figure 4.8–Supplemental Figure 4.12, P > 5 × 10
-8
). Quantile-quantile
plots did not reveal overdispersion of P values, nor did we observe inflation of the test
statistics (l ≤ 1.00, Supplemental Figure 4.13 and Supplemental Figure 4.14).
For each EBV antigen, we tested associations between genetic variation and
antibody level (continuous outcomes). This analysis was restricted to participants
exceeding the corresponding MFI threshold for antibody positivity. Results are described
in Table 4.9, Figure 4.3, and Supplemental Figure 4.15–Supplemental Figure 4.17. We
observed an association between levels of antibodies to EBNA-1 and two imputed
variants: rs140444865 at 3p26 (intronic, GRM7, P = 5.47 × 10
-9
) and rs115805790 at
2q35 (intergenic INHA; STK11IP, P = 2.70 × 10
-8
). A signal in an intergenic region at
10q25 (rs112173318 and rs112871292, intergenic SORCS1; RNU6−53P) also
approached genome-wide significance for association with EBNA-1 (P = 5.02 × 10
-8
). No

85
associations with antibodies to other EBV antigens surpassed the threshold for genome-
wide significance (P > 5 × 10
-8
). Quantile-quantile plots did not reveal overdispersion of
P values, nor did we observe inflation of the test statistics (l ≤ 1.02, Supplemental Figure
4.18).
Sensitivity analyses of the associations between rs115805790 and levels of
antibodies to EBNA-1 by participant characteristics are described in Table 4.10. The
association between rs115805790 and EBNA-1 is stronger among participants who were
not diagnosed with NSHL (P = 0.04) and those who did not report a positive history of
IM (P = 0.04). Data were insufficient to assess potential interaction with tumor EBV
status, and we did not observe effect modification by age at diagnosis, age at blood draw,
or time between diagnosis and blood draw (P ≥ 0.19). Sensitivity analysis for the
association between rs140444865 and MFI of antibodies to EBNA-1 did not indicate
interaction with HL histology (nodular sclerosing vs. other), tumor EBV status, self-
reported IM history, age at diagnosis, age at blood draw, or time between diagnosis and
blood draw (Pinteraction ≥ 0.14). There was weak evidence of interaction between
rs140444865 and sex on EBNA-1 (P = 0.10).

DISCUSSION
We observed a strong association between EBV viral load and variants at 3q25 in
the MUC4 gene among survivors of adolescent and young adult Hodgkin lymphoma. For

86
each A allele of 195489009, a missense MUC4 mutation, EBV viral load was roughly
1000 times higher (P = 7.22 × 10
-9
). Results were consistent across discovery sets.
Mucin 4 (MUC4), a transmembrane mucin protein encoded by the MUC4 gene, is
responsible for protecting and lubricating epithelial surfaces in the urogenital, digestive,
and respiratory tracts (194). Expression of MUC4 has been implicated in the pathogenesis
and prognosis for several malignancies including pancreatic, colon, and stomach cancers
(195). Mucins impact inflammation and immune response via direct interaction with
leukocytes and other immune cells. Transmembrane mucins such as MUC4 may hinder
leukocyte motility and activation status and prevent the approach of antigen-presenting
cells (196). It is known that mucins can influence immunotherapy response in melanoma
patients (197). Association between mucins and EBV has not been studied extensively,
but Kondo et al. found that EBV latent membrane protein 1 (LMP1) can induce MUC1
expression (198).
The preponderance of participants in our study population were of European
ethnicity/nationality, and our results are likely generalizable to other HL survivors of
European heritage. This European skew is consistent with the broader population
distribution of HL survivors. However, we cannot be sure our findings are generalizable
to other ethnic groups. Exploring the association in other populations (non-European HL
survivors, healthy European and non-European populations, and those who have other
EBV-related conditions and neoplasms such as infectious mononucleosis and specific
non-Hodgkin lymphoma subtypes) will allow us to further investigate the potential
relationship between MUC4 variation and EBV viral load.

87
We also observed associations between antibodies to EBNA-1 and variants in and
near GRM7, INHA, and STK11IP in the USC discovery set. Metabotropic glutamate
receptor 7 modulates neurotransmission, and polymorphisms in GRM7 are related to
neurodevelopmental disorders (199). The α subunit of inhibin (INHA) is a member of the
transforming growth factor beta (TGFβ) family, which regulates angiogenesis. INHA has
been implicated in tumor metastasis and vascularization of ovarian cancers (200).
Through its interaction with STK11, serine/threonine kinase 11 interacting protein
(STK11IP, alias LIP1) regulates STK11 function (cellular proliferation, signaling, and
apoptosis (201,202).
In this study, we identified associations between genetic variants, EBV viral load,
and antibodies to EBV among HL survivors. Variants in and near MUC4, GRM7, INHA,
and STK1 are associated with higher EBV viral load and higher levels of antibodies to
EBNA-1 among Hodgkin lymphoma survivors. Targeted studies to explore functional
consequences and validate associations in broader population groups are warranted.

88
TABLES
Table 4.1. Characteristics of Participants from USC and UPRD Discovery Sets
USC (N=166) UPRD (N=106)

N % N %
Sex Female 68 41%

43 41%

Male 98 59%

63 59%

Histology Classic (NOS) 9 5%

0 0%

Mixed cellularity 16 10%

9 8%

Nodular sclerosis 131 79%

75 71%

Other/NOS 10 6%

22 21%

Tumor EBV
status
EBV negative 82 49%

49 46%
EBV positive 15 9%

20 19%
Unknown 69 42%

37 35%

Infectious
mononucleosis
history
IM not reported 132 80%

93 46%
IM reported 34 20%

11 19%
Unknown 0 0%

2 35%

Ethnicity
a
European 108 65%

106 100%
Hispanic or Latinx 1 1%

0 0%
Middle Eastern 5 3%

0 0%
Unknown 52 31%

0 0%

N Mean (SD) N Mean (SD)
Age at diagnosis 166 29.6 (8.5)

105 26.5 (7.0)
Age at blood draw 149 36.5 (10.6)

106 32.1 (8.8)
Time between diagnosis and blood draw (years) 149 6.5 (8.1)

105 5.7 (5.8)
EBV viral load 166 192 (1401)

106 106 (336)
EA-D (MFI) 143 2670 (2893)

EBNA (MFI) 143 5621 (4292)

VCA p18 (MFI) 143 8457 (3891)

ZEBRA (MFI) 143 1796 (1979)

EA-D: early antigen-diffuse; EBNA-1: Epstein-Barr virus nuclear antigen 1; EBV: Epstein-Barr virus; IM: infectious
mononucleosis; MFI: median fluorescence intensity; NOS: not otherwise specified; SD: standard deviation; UPRD:
Université de Paris René Descartes; USC: University of Southern California; VCA: viral capsid antigen; ZEBRA: Z-
Epstein-Barr virus replication activator.
a
Ethnicity was self-reported by USC participants. French legal restrictions prevented UPRD investigators from
collecting self-reported ethnicity data, and all participants were assumed to be of European ethnicity. The European
category includes Turkish and European American. Middle Eastern includes Iranian.

89

Table 4.2. Pre- and Post-Imputation Availability of Genotyping Data for Study Participants
USC (N)

UPRD (N)

Meta-analysis (N)

Participants Variants Participants Variants Participants Variants
Genotyped

366 566,330

448 242,015

--

Passed pre-imputation QC
a

366 531,757

448 241,034

Passed post-imputation QC
b
Viral load 166 7,685,623

106 7,250,996

272 6,776,403

Antibodies: dichotomized outcomes
c
143 7,783,313

Antibodies: EA-D MFI
d
120 7,783,313

Antibodies: EBNA-1 MFI
d
124 7,783,313

Antibodies: VCA p18 MFI
d
131 7,783,313

Antibodies: ZEBRA MFI
d
122 7,783,313
EA-D: early antigen-diffuse; EBNA-1: Epstein-Barr virus nuclear antigen 1; MFI: median fluorescence intensity; QC: quality control; UPRD: Université de Paris René Descartes;
USC: University of Southern California; VCA: viral capsid antigen; ZEBRA: Z-Epstein-Barr virus replication activator.

a
Participant exclusion criterion: call rate < 95%. Variant exclusion criteria: monomorphic, minor allele frequency (MAF) < 0.01, call rate < 95%, or strong deviation from Hardy-
Weinberg equilibrium (P < 1 × 10
-5
).

b
Participant exclusion criteria: saliva (rather than blood) sample submitted for genotyping or missing phenotype. Variant exclusion criteria: imputation quality < 0.3, minor allele
frequency (MAF) < 0.05 among participants with non-missing phenotype, or strong deviation from Hardy-Weinberg equilibrium (P < 1 × 10-5) . Meta-analysis variant exclusion
criterion: missing in USC or UPRD data.

c
Serology dichotomized outcomes: EBV, EA-D, EBNA-1, VCA p18, and ZEBRA.

d
Analysis for each protein is restricted to participants with MFI value at or above corresponding threshold.

90

Table 4.3. Associations Between Chromosome 3 Variants and EBV Viral Load Among 166 Participants with Hodgkin Lymphoma in
USC Discovery Set
Variant rs ID
a
Position (BP)
a
Ref Alt Function
b
Gene
b

Imputation
quality (r
2
) MAF Estimate
c
95% CI
c
P value
c

rs2246901 195489009 C A Missense MUC4 Genotyped 0.30 1,349 83, 21878 1.06 × 10
-6

rs842220 195491423 G A Intronic MUC4 0.94 0.31 1,349 83, 21878 1.06 × 10
-6

rs2550263 195489669 C T Intronic MUC4 0.94 0.31 1,208 73, 19861 1.71 × 10
-6

rs2550262 195490144 C A Intronic MUC4 0.94 0.31 1,208 73, 19861 1.71 × 10
-6

rs2246771 195487737 A G Intronic MUC4 0.94 0.30 1,096 67, 17906 2.17 × 10
-6

rs6781229 195552216 T C Intergenic MUC4;LINC01983 0.82 0.26 899 46, 17783 1.47 × 10
-5

rs844518 195492817 C T Intronic MUC4 Genotyped 0.28 830 51, 13614 5.28 × 10
-6

rs2688525 195485231 G A Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

rs2641719 195485439 C T Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

rs2641718 195485732 G A Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

rs2550272 195486633 A C Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

rs2688526 195486680 T C Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

rs2550273 195486705 G A Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

rs2550275 195486869 G A Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

rs2641717 195487117 T C Intronic MUC4 0.94 0.31 979 59, 16218 3.45 × 10
-6

Alt: alternate allele; CI: confidence interval; EBV: Epstein-Barr virus; MAF: minor allele frequency; Ref: reference allele; rs ID: Reference SNP cluster ID; USC: University of
Southern California.

a
Chromosome location based on National Center for Biotechnology Information Human Genome Build 37 coordinates.

b
Functional annotation conducted with ANNOVAR using the Homo Sapiens hg19 genome assembly (hg19/GRCh37) and the Reference Sequence (RefSeq) database for gene
definition.

c
!s, CIs, and P values in discovery sets calculated using linear regression models on log10 transformed EBV viral load adjusted for sex and the first ten principal components. !s
and CIs were exponentiated, and the presented estimate should be interpreted as the proportional change in the expected geometric mean of EBV viral load for each minor allele.

91

Table 4.4. Associations Between Chromosome 3 Variants and EBV Viral Load Among 106 Participants with Hodgkin Lymphoma in
UPRD Discovery Set
Variant rs ID
a
Position (BP)
a
Ref Alt Function
b
Gene
b

Imputation
quality (r
2
) MAF Estimate
c
95% CI
c
P value
c

rs2246901 195489009 C A Missense MUC4 0.83 0.30 518 7, 40644 5.97 × 10
-3

rs842220 195491423 G A Intronic MUC4 0.83 0.30 518 7, 40644 5.97 × 10
-3

rs2550263 195489669 C T Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

rs2550262 195490144 C A Intronic MUC4 0.83 0.30 518 7, 40644 5.97 × 10
-3

rs2246771 195487737 A G Intronic MUC4 0.83 0.30 710 8, 62806 4.96 × 10
-3

rs6781229 195552216 T C Intergenic MUC4;LINC01983 0.77 0.23 2,553 32, 202302 6.51 × 10
-4

rs844518 195492817 C T Intronic MUC4 0.82 0.29 977 11, 88105 3.40 × 10
-3

rs2688525 195485231 G A Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

rs2641719 195485439 C T Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

rs2641718 195485732 G A Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

rs2550272 195486633 A C Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

rs2688526 195486680 T C Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

rs2550273 195486705 G A Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

rs2550275 195486869 G A Intronic MUC4 Genotyped 0.27 518 7, 40644 5.97 × 10
-3

rs2641717 195487117 T C Intronic MUC4 0.84 0.30 518 7, 40644 5.97 × 10
-3

Alt: alternate allele; CI: confidence interval; EBV: Epstein-Barr virus; MAF: minor allele frequency; Ref: reference allele; rs ID: Reference SNP cluster ID; UPRD: Université de Paris
René Descartes.

a
Chromosome location based on National Center for Biotechnology Information Human Genome Build 37 coordinates.

b
Functional annotation conducted with ANNOVAR using the Homo Sapiens hg19 genome assembly (hg19/GRCh37) and the Reference Sequence (RefSeq) database for gene
definition.

c
!s, CIs, and P values in discovery sets calculated using linear regression models on log10 transformed EBV viral load adjusted for sex and the first ten principal components. !s
and CIs were exponentiated, and the presented estimate should be interpreted as the proportional change in the expected geometric mean of EBV viral load for each minor allele.

92

Table 4.5. Associations Between Chromosome 3 Variants and EBV Viral Load Among 272 Participants with Hodgkin Lymphoma in
Meta-Analysis
Variant rs ID
a
Position (BP)
a
Ref Alt Function
b
Gene
b
Estimate 95% CI P value I
2
Phom
rs2246901 195489009 C A Missense MUC4 1,022 473, 1572 7.22 × 10
-9
** 0% 0.72
rs842220 195491423 G A Intronic MUC4 1,022 473, 1572 7.22 × 10
-9
** 0% 0.72
rs2550263 195489669 C T Intronic MUC4 943 427, 1459 1.23 × 10
-8
** 0% 0.75
rs2550262 195490144 C A Intronic MUC4 943 427, 1459 1.23 × 10
-8
** 0% 0.75
rs2246771 195487737 A G Intronic MUC4 971 439, 1503 1.29 × 10
-8
** 0% 0.87
rs6781229 195552216 T C Intergenic MUC4;LINC01983 1,253 564, 1941 1.41 × 10
-8
** 0% 0.70
rs844518 195492817 C T Intronic MUC4 869 383, 1354 2.36 × 10
-8
** 0% 0.95
rs2688525 195485231 G A Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
rs2641719 195485439 C T Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
rs2641718 195485732 G A Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
rs2550272 195486633 A C Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
rs2688526 195486680 T C Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
rs2550273 195486705 G A Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
rs2550275 195486869 G A Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
rs2641717 195487117 T C Intronic MUC4 813 356, 1269 2.67 × 10
-8
** 0% 0.81
Alt: alternate allele; CI: confidence interval; EBV: Epstein-Barr virus; MAF: minor allele frequency; Ref: reference allele; rs ID: Reference SNP cluster ID; UPRD: Université de Paris
René Descartes; USC: University of Southern California.

** P value < 5 × 10
-8
.

a
Chromosome location based on National Center for Biotechnology Information Human Genome Build 37 coordinates. All listed variants were imputed with the exception of
rs2246901 (genotyped in USC set), rs844518 (genotyped in USC set), and rs2550275 (genotyped in UPRD set).

b
Functional annotation conducted with ANNOVAR using the Homo Sapiens hg19 genome assembly (hg19/GRCh37) and the Reference Sequence (RefSeq) database for gene
definition.

c
Meta-analysis !s and P values calculated using inverse variance-weighted fixed-effects models!s were exponentiated, and the presented estimate should be interpreted as the
proportional change in the expected geometric mean of EBV viral load for each minor allele. Homogeneity P value (Phom) calculated using Cochran's Q statistic.

93
Table 4.6. ANNOVAR Functional Annotation of rs2246901
Gene MUC4
Protein change

p.Ala4821Ser

CADD Phred

22.9

ExAC frequency

0.6823

Mutation significance predictions
FATHMM Tolerated

LRT Neutral

MetaLR Tolerated

MetaSVM Tolerated

Mutation Assessor Functional (Medium) *

MutationTaster Polymorphism Automatic

PolyPhen 2 HDIV Probably Damaging *

PolyPhen 2 HVAR Probably Damaging *

PROVEAN Neutral

SIFT Tolerated
CADD: combined annotation dependent depletion; ExAC: exome aggregation consortium; FATHMM: functional
analysis through hidden Markov models; LRT: likelihood ratio test; MetaLR: meta-analytic logistic regression;
MetaSVM: meta-analytic support vector machine; PolyPhen 2: polymorphism phenotyping v2; PROVEAN: protein
variation effect analyzer; SIFT: sorting intolerant from tolerant.

94

Table 4.7. Association Between the A Allele of rs2246901 and EBV Viral Load Among 272 Hodgkin Lymphoma Survivors Stratified by
Discovery Set, Histology, Tumor EBV Status, and Infectious Mononucleosis History
N Mean (SD) Estimate
a
95% CI
a
P value
a

Discovery set USC 166 192 (1401) 621 37, 10363 <0.001 *

UPRD 106 106 (336) 695 5, 104641 0.01 *

Interaction

0.73

Histology Nodular sclerosis 206 173 (1263) 655 40, 10714 <0.001 *

MC/other/NOS 66 113 (374) 2 0, 1100 0.86

Interaction

0.60

Tumor EBV status
b

EBV negative 131 27 (84) 152 4, 6730 0.01 *
EBV positive 35 120 (296) 680 0, >10000000 0.29

Unknown 106 334 (1764) 5084 103, 245088 <0.001 *
Interaction

0.73

Infectious mononucleosis history
b

IM not reported 225 110 (714) 438 28, 6504 <0.001 *
IM reported 45 400 (2231) 6454 50, 872652 <0.001 *
Unknown 2 127 (25) --
c

Interaction

0.20

CI: confidence interval; EBV: Epstein-Barr virus; IM: infectious mononucleosis; SD: standard deviation; UPRD: Université de Paris René Descartes; USC: University of Southern
California.
* P value < 0.05.
a
!s, CIs, and P values calculated using linear regression models on log transformed EBV viral load adjusted for the first ten principal components. With the exception of the sex-
stratified model, all models are adjusted for sex. !s and CIs were exponentiated, and the presented estimate should be interpreted as the proportional change in the expected
geometric mean of EBV viral load for each A allele of rs2246901.
b
Interaction P values were calculated on a continuous scale using models that excluded missing/unknown data.
c
Insufficient data to compute effect estimate.

95

Table 4.8. Association Between the A Allele of rs2246901 and EBV Viral Load Among 272 Hodgkin Lymphoma Survivors Stratified by
Sex, Age at Diagnosis, Age at Blood Draw, and Time between Diagnosis and Blood Draw
N Mean (SD) Estimate
a
95% CI
a
P value
a

Sex Female 111 41 (99) 162 4, 6419 0.01 *

Male 161 239 (1442) 1767 60, 51664 <0.001 *

Interaction

0.25

Age at diagnosis (years)
b
≤ 20
54 214 (1332) 2531 2, 1806087 0.03 *
21–30
122 94 (355) 232 4, 12637 0.01 *
> 30
95 209 (1551) 1295 21, 85006 <0.001 *
Unknown
1 203

--
c

Interaction

0.77

Age at blood draw (years)
b
≤ 40
193 59 (252) 213 13, 4791 <0.001
*
41–45
34 304 (1679) 432200 160, >10000000 0.01
*
> 45
28 151 (490) 345 0, 1188279 0.19

Unknown
17 1008 (3620) 76 0, >10000000 0.64

Interaction

0.23

Time from diagnosis to blood draw (years)
b
≤ 15
232 54 (236) 564 38, 8472 <0.001
*
> 15
22 597 (2124) 2309 0, >10000000 0.17

Unknown
18 963 (3517) 9 0, >10000000 0.79

Interaction

0.20

CI: confidence interval; EBV: Epstein-Barr virus; IM: infectious mononucleosis; SD: standard deviation.
* P value < 0.05.
a
!s, CIs, and P values calculated using linear regression models on log transformed EBV viral load adjusted for the first ten principal components. With the exception of the sex-
stratified model, all models are adjusted for sex. !s and CIs were exponentiated, and the presented estimate should be interpreted as the proportional change in the expected
geometric mean of EBV viral load for each A allele of rs2246901.
b
Interaction P values were calculated on a continuous scale using models that excluded missing/unknown data.
c
Insufficient data to compute effect estimate.

96

Table 4.9. Variants Most Strongly Associated with MFI Levels of Antibodies to EBV Antigens Among USC Participants Exceeding
MFI Threshold for Antibody Positivity
Antigen Variant rs ID
a
Chr Position (BP)
a
Ref Alt Function
b
Gene
b

Imputation
quality (r
2
) MAF Estimate
c
95% CI
c
P
c

EA-D rs4395181 18 64801045 C T Intergenic MIR5011;DSEL 0.96 0.39 0.45 0.34, 0.59 6.63 × 10
-8

EA-D rs9964036 18 64790088 A C Intergenic MIR5011;DSEL 0.97 0.39 0.45 0.34, 0.59 1.87 × 10
-7

EBNA-1 rs140444865 3 7610293 G A Intronic GRM7 0.77 0.06 0.29 0.2, 0.43 5.47 × 10
-9
**
EBNA-1 rs115805790 2 220453909 A C Intergenic INHA;STK11IP 0.79 0.06 0.27 0.17, 0.41 2.70 × 10
-8
**
VCAp18 rs11594715 10 107485065 A G Intronic (ncRNA) LOC101927549 0.97 0.08 0.63 0.53, 0.74 1.56 × 10
-7

VCAp18 rs117726444 10 107488975 T C Intronic (ncRNA) LOC101927549 0.97 0.08 0.63 0.53, 0.74 1.56 × 10
-7

ZEBRA rs7168418 15 84655733 C T Intronic ADAMTSL3 0.98 0.23 1.95 1.49, 2.54 2.88 × 10
-6

Zebra rs6603013 15 84662420 G A Intronic ADAMTSL3 0.98 0.23 1.95 1.49, 2.54 2.88 × 10
-6

Alt: alternate allele; BP: base pairs; CI: confidence interval; EA-D: early antigen-diffuse; EBNA-1: Epstein-Barr virus nuclear antigen 1; EBV: Epstein-Barr virus; MAF: minor allele
frequency; MFI: median fluorescence intensity; Ref: reference allele; rs ID: Reference SNP cluster ID; USC: University of Southern California; VCA: viral capsid antigen; ZEBRA: Z-
Epstein-Barr virus replication activator.
** P value < 5 × 10
-8
.
a Chromosome location based on National Center for Biotechnology Information Human Genome Build 37 coordinates. All listed variants were imputed.
b
Annotation conducted with ANNOVAR using the Homo Sapiens hg19 genome assembly (hg19/GRCh37) and the Reference Sequence (RefSeq) database for gene definition.
c !s, CIs, and P values calculated using linear regression models on log transformed antibody levels (MFI) adjusted for sex and the first ten principal components. !s and CIs were
exponentiated, and the presented estimate should be interpreted as the proportional change in the expected geometric mean of antibody MFI for each minor allele.

97
Table 4.10. Association Between the C Allele of rs115805790 and Levels of Antibodies to
EBNA-1 Antigen Among 124 USC Participants with EBNA-1 Levels Surpassing 411 MFI
by Sex, Histology, Tumor EBV Status, and Self-Reported Infectious Mononucleosis
History
N Mean (SD) Estimate
a
95% CI
a
P value
a

Sex Female 46 6224 (3831) 0.20 0.09, 0.47 <0.001 *

Male 78 6606 (4089) 0.27 0.15, 0.49 <0.001 *

Interaction

0.55

Histology Nodular sclerosis 97 6183 (3527) 0.30 0.19, 0.47 <0.001 *

MC/other/NOS 27 7475 (5272) 0.07 0.01, 0.47 0.01 *

Interaction

0.04 *

Tumor EBV
status
EBV negative 71 6202 (3826) 0.29 0.16, 0.52 <0.001 *
EBV positive 11 6446 (5323) --
c

Unknown 42 6913 (3922) 0.40 0.13, 1.21 0.11

Interaction

0.11

Infectious
mononucleos
is history
IM not reported 99 6886 (3971) 0.30 0.18, 0.51 <0.001 *
IM reported 25 4794 (3648) 0.28 0.1, 0.66 0.01 *
Interaction

0.04 *

Age at
diagnosis
(years)
b

≤ 20
17 7,117 (2867) 0.58 0.17, 1.33 0.25

21–30
50 6,553 (4330) 0.14 0.07, 0.26 <0.001 *
> 30
57 6,191 (3987) 0.37 0.18, 0.76 0.01 *
Interaction

0.47

Age at blood
draw (years)
b

≤ 40
79 6,456 (3902) 0.30 0.17, 0.52 <0.001
*
41–45
27 7,427 (4056) 0.51 0.15, 1.75 0.28

> 45
18 5,056 (4029) 0.39 0.09, 1.64 0.20

Interaction

0.19

Time from dx
to blood draw
(years)
b

≤ 15
109 6,647 (4066) 0.29 0.18, 0.46 <0.001
*
> 15
15 5,135 (3127) 0.34 0.01, 13.95 0.56

Interaction

0.20

CI: confidence interval; EBNA-1: Epstein-Barr Virus Nuclear Antigen 1; EBV: Epstein-Barr virus; IM: infectious
mononucleosis; MFI: Median Fluorescence Intensity; SD: standard deviation; USC: University of Southern California.
* P value < 0.05.
a
!s, CIs, and P values calculated using linear regression models on log transformed EBNA-1 MFI adjusted for the first
ten principal components. With the exception of the sex-stratified model, all models are adjusted for sex. !s and CIs
were exponentiated, and the presented estimate should be interpreted as the proportional change in the expected
geometric mean of EBNA-1 MFI for each C allele of rs115805790.
b
Interaction P values for age at diagnosis, age at blood draw, and time from diagnosis to blood draw were calculated
on a continuous scale using models that excluded missing/unknown data.
c
Insufficient sample size to estimate effect.

98
Table 4.11. Association Between the A Allele of rs140444865 and Levels of Antibodies to
EBNA-1 Antigen Among 124 USC Participants with EBNA-1 Levels Surpassing 411 MFI
by Sex, Histology, Tumor EBV Status, and Self-Reported Infectious Mononucleosis
History
N Mean (SD) Estimate
a
95% CI
a
P value
a

Sex Female 46 6224 (3831) 0.25 0.15, 0.42 <0.001 *

Male 78 6606 (4089) 0.48 0.26, 0.89 0.02 *

Interaction

0.10

Histology Nodular sclerosis 97 6183 (3527) 0.41 0.28, 0.60 <0.001 *

MC/other/NOS 27 7475 (5272) 0.19 0.02, 2.36 0.19

Interaction

0.14

Tumor EBV
status
EBV negative 71 6202 (3826) 0.39 0.24, 0.60 <0.001 *
EBV positive 11 6446 (5323) --
c

Unknown 42 6913 (3922) 1.43 0.24, 7.38 0.69

Interaction

Infectious
mononucleosis
history
IM not reported 99 6886 (3971) 0.55 0.32, 0.96 0.04 *
IM reported 25 4794 (3648) 0.29 0.16, 0.46 <0.001 *
Interaction

0.14

Age at
diagnosis
(years)
b

≤ 20
17 7,117 (2867) 0.58 0.17, 1.33 0.25

21–30
50 6,553 (4330) 0.44 0.24, 0.80 0.01 *
> 30
57 6,191 (3987) 0.36 0.19, 0.69 <0.001 *
Interaction

0.43

Age at blood
draw (years)
b

≤ 40
79 6,456 (3902) 0.41 0.25, 0.69 <0.001
*
41–45
27 7,427 (4056) 0.48 0.18, 1.28 0.15

> 45
18 5,056 (4029) 0.56 0.25, 1.36 0.20

Interaction

0.67

Time from dx to
blood draw
(years)
b

≤ 15 10
9 6,647 (4066) 0.37 0.24, 0.59 <0.001
*
> 15
15 5,135 (3127) 0.32 0.07, 5.14 0.53

Interaction

0.88

CI: confidence interval; EBNA-1: Epstein-Barr Virus Nuclear Antigen 1; EBV: Epstein-Barr virus; IM: infectious
mononucleosis; MFI: Median Fluorescence Intensity; SD: standard deviation; USC: University of Southern California.
* P value < 0.05.
a
!s, CIs, and P values calculated using linear regression models on log transformed EBNA-1 MFI adjusted for the first
ten principal components. With the exception of the sex-stratified model, all models are adjusted for sex. !s and CIs
were exponentiated, and the presented estimate should be interpreted as the proportional change in the expected
geometric mean of EBNA-1 MFI for each A allele of rs140444865.
b
Interaction P values for age at diagnosis, age at blood draw, and time from diagnosis to blood draw were calculated
on a continuous scale using models that excluded missing/unknown data.
c
Insufficient sample size to estimate effect.

99
FIGURES
Figure 4.1. Genome-Wide Association Manhattan Plot for EBV Viral Load: Results from
Meta-Analysis

The genome-wide association meta-analysis of EBV viral load includes data from two
study sites, 273 participants, and approximately 6.8 million variants after imputation. The
red line indicates genome-wide significant associations (P = 5 × 10
-8
). The blue line
indicates suggestive associations that may warrant further investigation (P = 1 × 10
-5
).

100
Figure 4.2. LocusZoom Plot for Associations Between MUC4 Variants in Chromosome 3
and EBV Viral Load: Results from Meta-Analysis

The genome-wide association meta-analysis of EBV viral load includes data from two
study sites, 273 participants, and approximately 6.8 million variants after imputation. The
color-coding represents the r
2
measure of linkage disequilibrium between rs2246901 and
other variants. Variants in highest linkage disequilibrium with rs2246901 are shown in
red.

101
Figure 4.3. Genome-Wide Association Manhattan Plot for Median Fluorescence Intensity
(MFI) of Antibodies to Epstein-Barr Virus Nuclear Antigen 1 (EBNA-1) Among
Participants Exceeding MFI Threshold for EBNA-1 Positivity: Results from University of
Southern California Discovery Set

The genome-wide association study of EBNA-1 MFI includes data from 124 University
of Southern California participants exceeding a 411 EBNA-1 MFI threshold and
approximately 7.8 million variants after imputation. The red line indicates genome-wide
significant associations (P = 5 × 10
-8
). The blue line indicates suggestive associations that
may warrant further investigation (P = 1 × 10
-5
).

102
Figure 4.4. LocusZoom Plot for Association Between rs115805790 (INHA;STK11IP) in
Chromosome 2 and Median Fluorescence Intensity (MFI) of Antibodies to Epstein-Barr
Virus Nuclear Antigen 1 (EBNA-1) Among Participants Exceeding MFI Threshold for
EBNA-1 Positivity: Results from University of Southern California Discovery Set

The genome-wide association study of EBNA-1 MFI includes data from 124 University
of Southern California participants exceeding a 411 EBNA-1 MFI threshold and
approximately 7.8 million variants after imputation. The color-coding represents the r
2

measure of linkage disequilibrium between rs115805790 and other variants. Variants in
highest linkage disequilibrium with rs115805790 are shown in red.

103
Figure 4.5. LocusZoom Plot for Association Between rs140444865 (GRM7) in
Chromosome 3 and Median Fluorescence Intensity (MFI) of Antibodies to Epstein-Barr
Virus Nuclear Antigen 1 (EBNA-1) Among Participants Exceeding MFI Threshold for
EBNA-1 Positivity: Results from University of Southern California Discovery Set

The genome-wide association study of EBNA-1 MFI includes data from 124 University
of Southern California participants exceeding a 411 EBNA-1 MFI threshold and
approximately 7.8 million variants after imputation. The color-coding represents the r
2

measure of linkage disequilibrium between rs140444865 and other variants. Variants in
highest linkage disequilibrium with rs140444865 are shown in red.

104
SUPPLEMENTAL MATERIALS
Supplemental Table 4.1. Characteristics of 143 Hodgkin Lymphoma Survivors from USC
with Sufficient Blood Sample for Analysis of Antibodies to EBV Antigens
N %
Sex
Female 54 38%

Male 89 62%

Histology
Classic Hodgkin 9 6%

Mixed cellularity 13 9%

Nodular lymphocyte-predominant 1 1%

Nodular sclerosing 113 79%

Unknown 7 5%

Tumor EBV status
EBV negative 80 56%

EBV positive 14 10%

Unknown 49 34%

Infectious mononucleosis history
IM not reported 97 68%

IM reported 25 17%

Unknown 21 15%

Self-reported ethnicity
European (includes Turkish or American) 103 72%

Hispanic or Latinx 1 1%

Middle Eastern (includes Iranian) 5 3%

Unknown 34 24%

N Mean (SD)
Age at Diagnosis 143 29.8 (8.0)
Age at Blood Draw 143 36.3 (10.5)
EBV viral load
143 103 (846)
EA-D (MFI)
143 2670 (2893)
EBNA (MFI)
143 5621 (4292)
VCA p18 (MFI)
143 8457 (3891)
ZEBRA (MFI)
143 1796 (1979)
EA-D: early antigen-diffuse; EBNA-1: Epstein-Barr virus nuclear antigen 1; EBV: Epstein-Barr virus; IM: infectious
mononucleosis; MFI: median fluorescence intensity; SD: standard deviation; USC: University of Southern California;
VCA: viral capsid antigen; ZEBRA: Z-Epstein-Barr virus replication activator.

105
Supplemental Figure 4.1. Distribution of EBV Viral Copy number by Participant
Characteristics
Discovery set
P = 0.03

Tumor EBV status
P = 0.03

IM history
P = 0.87

Histology
P = 0.19

Sex
P = 0.74

Time from diagnosis to blood draw
P = 0.45

Age at diagnosis
P = 0.59

Age at blood draw
P = 0.54

Kruskal-Wallis and Wilcoxon rank sum tests were used to compare EBV viral copy
number across participant characteristics. Whiskers on plots indicate 1.5 × interquartile
range (IQR). EBV viral copy number varied across discovery sets in our study
population.

106
Supplemental Figure 4.2. Quantile-Quantile Plots (-log10 Scale) for EBV Viral Load:
Results from University of Southern California and Université de Paris René Descartes
Discovery Sets
University of Southern California Discovery Set
l = 1.03
Université de Paris René Descartes Discovery Set
l = 1.02

The genome-wide association study of EBV viral load for the University of Southern
California discovery set includes data from 166 participants and approximately 7.7
million variants after imputation. The genome-wide association study of EBV viral load
for the Université de Paris René Descartes discovery set includes data from 107
participants and approximately 7.3 million variants after imputation. We did not observe
overdispersion of P values or inflation of the test statistic.

107
Supplemental Figure 4.3. Quantile-Quantile Plot (-log10 Scale) for EBV Viral Load:
Results from Meta-Analysis
Meta-analysis
l = 1.03

The genome-wide association meta-analysis of EBV viral load includes data from two
study sites, 273 participants, and approximately 6.8 million variants after imputation. We
did not observe overdispersion of P values or inflation of the test statistic.

108
Supplemental Figure 4.4. Distribution of Median Fluorescence Intensity (MFI) of
Antibodies to Early Antigen-Diffuse (EA-D) by Participant Characteristics
Discovery set

Tumor EBV status
P = 0.66

IM history
P = 0.03

Histology
P = 0.04

Sex
P = 0.38

Time from diagnosis to blood draw
P = 0.76

Age at diagnosis
P > 0.99

Age at blood draw
P = 0.09

Kruskal-Wallis and Wilcoxon rank sum tests were used to compare EA-D MFI across
participant characteristics. Whiskers on plots indicate 1.5 × interquartile range (IQR). The
distribution of EA-D MFI varied by IM history and histology.

109
Supplemental Figure 4.5. Distribution of Median Fluorescence Intensity (MFI) of
Antibodies Epstein-Barr Virus Nuclear Antigen 1 (EBNA-1) by Participant
Characteristics
Discovery set

Tumor EBV status
P = 0.50

IM history
P = 0.08

Histology
P = 0.08

Sex
P = 0.58

Time from diagnosis to blood draw
P = 0.31

Age at diagnosis
P = 0.72

Age at blood draw
P = 0.22

Kruskal-Wallis and Wilcoxon rank sum tests were used to compare EBNA-1 MFI across
participant characteristics. Whiskers on plots indicate 1.5 × interquartile range (IQR).

110
Supplemental Figure 4.6. Distribution of Median Fluorescence Intensity (MFI) of
Antibodies Viral Capsid Antigen p18 (VCA p18) by Participant Characteristics
Discovery set

Tumor EBV status
P = 0.06

IM history
P = 0.28

Histology
P = 0.16

Sex
P = 0.58

Time from diagnosis to blood draw
P = 0.30

Age at diagnosis
P = 0.23

Age at blood draw
P = 0.81

Kruskal-Wallis and Wilcoxon rank sum tests were used to compare VCA p18 MFI across
participant characteristics. Whiskers on plots indicate 1.5 × interquartile range (IQR).

111
Supplemental Figure 4.7. Distribution of Median Fluorescence Intensity (MFI) of
Antibodies Z-Epstein-Barr Virus Replication Activator (ZEBRA) by Participant
Characteristics
Discovery set

Tumor EBV status
P = 0.36

IM history
P = 0.16

Histology
P = 0.41

Sex
P = 0.55

Time from diagnosis to blood draw
P = 0.67

Age at diagnosis
P = 0.15

Age at blood draw
P = 0.04

Kruskal-Wallis and Wilcoxon rank sum tests were used to compare ZEBRA MFI across
participant characteristics. Whiskers on plots indicate 1.5 × interquartile range (IQR). The
distribution of ZEBRA MFI varied by IM history and histology

112
Supplemental Figure 4.8. Genome-Wide Association Manhattan Plot for Presence of
Antibodies to Early Antigen-Diffuse (EA-D): Results from University of Southern
California Discovery Set

The genome-wide association study of the presence of antibodies to EA-D includes data
from 143 University of Southern California participants and approximately 7.8 million
variants after imputation. Median fluorescence intensity (MFI) levels exceeding a 110
MFI threshold were considered positive. The blue line indicates suggestive associations
that may warrant further investigation (P = 1 × 10
-5
). No associations reached genome-
wide significance (P = 5 × 10
-8
).

113
Supplemental Figure 4.9. Genome-Wide Association Manhattan Plot for the Presence of
Antibodies to Epstein-Barr Virus Nuclear Antigen 1 (EBNA-1): Results from University
of Southern California Discovery Set

The genome-wide association study of the presence of antibodies to EBNA-1 includes
data from 143 University of Southern California participants and approximately 7.8
million variants after imputation. Median fluorescence intensity (MFI) levels exceeding a
411 MFI threshold were considered positive. The blue line indicates suggestive
associations that may warrant further investigation (P = 1 × 10
-5
). No associations
reached genome-wide significance (P = 5 × 10
-8
).

114
Supplemental Figure 4.10. Genome-Wide Association Manhattan Plot for the Presence of
Antibodies to Viral Capsid Antigen p18 (VCA p18): Results from University of Southern
California Discovery Set

The genome-wide association study of the presence of antibodies to VCA p18 includes
data from 143 University of Southern California participants and approximately 7.8
million variants after imputation. Median fluorescence intensity (MFI) levels exceeding a
2526 MFI threshold were considered positive. The blue line indicates suggestive
associations that may warrant further investigation (P = 1 × 10
-5
). No associations
reached genome-wide significance (P = 5 × 10
-8
).

115
Supplemental Figure 4.11. Genome-Wide Association Manhattan Plot for the Presence of
Antibodies to Z-Epstein-Barr Virus Replication Activator (ZEBRA): Results from
University of Southern California Discovery Set

The genome-wide association study of the presence of antibodies to ZEBRA includes
data from 143 University of Southern California participants and approximately 7.8
million variants after imputation. Median fluorescence intensity (MFI) levels exceeding a
74 MFI threshold were considered positive. The blue line indicates suggestive
associations that may warrant further investigation (P = 1 × 10
-5
). No associations
reached genome-wide significance (P = 5 × 10
-8
).

116
Supplemental Figure 4.12. Genome-Wide Association Manhattan Plot for Serum EBV
Positivity: Results from University of Southern California Discovery Set

The genome-wide association study of serum Epstein-Barr virus (EBV) positivity
includes data from 143 University of Southern California participants and approximately
7.8 million variants after imputation. Participants were then classified as EBV positive if
the median fluorescence intensity for antibodies to two or more EBV proteins (early
antigen-diffuse, EBV nuclear antigen 1, viral capsid antigen p18, or Z-Epstein-Barr virus
replication activator) met or exceeded corresponding MFI thresholds. The blue line
indicates suggestive associations that may warrant further investigation (P = 1 × 10
-5
). No
associations reached genome-wide significance (P = 5 × 10
-8
).

117
Supplemental Figure 4.13. Quantile-Quantile Plots(-log10 Scale) for Presence of
Antibodies to EBV Antigens: Results from University of Southern California Discovery
Set

Early antigen-diffuse (EA-D)
l = 0.93

EBV nuclear antigen 1 (EBNA-1)
l = 0.96

Viral capsid antigen p18 (VCA p18)
l = 1.00

Z-Epstein-Barr virus replication activator (ZEBRA)
l = 0.96

The genome-wide association studies of the presence of antibodies to EBV antigens
include data from 143 University of Southern California participants and approximately
7.8 million variants after imputation. We did not observe overdispersion of P values or
inflation of the test statistic.

118
Supplemental Figure 4.14. Quantile-Quantile Plot (-log10 Scale) for Serum EBV
Positivity: Results from University of Southern California Discovery Set

l = 1.00

The genome-wide association study of serum Epstein-Barr virus (EBV) positivity
includes data from 143 University of Southern California participants and approximately
7.8 million variants after imputation. Participants were then classified as EBV positive if
the median fluorescence intensity for antibodies to two or more EBV proteins (early
antigen-diffuse, EBV nuclear antigen 1, viral capsid antigen p18, or Z-Epstein-Barr virus
replication activator) met or exceeded corresponding MFI thresholds. We did not observe
overdispersion of P values or inflation of the test statistic.

119
Supplemental Figure 4.15. Genome-Wide Association Manhattan Plot for Median
Fluorescence Intensity (MFI) of Antibodies to Early Antigen-Diffuse (EA-D) Among
Participants Exceeding MFI Threshold for EA-D Antibody Positivity: Results from
University of Southern California Discovery Set

The genome-wide association study of EA-D MFI includes data from 120 University of
Southern California participants exceeding a 110 EA-D MFI threshold and approximately
7.8 million variants after imputation. The red line indicates genome-wide significant
associations (P = 5 × 10
-8
). The blue line indicates suggestive associations that may
warrant further investigation (P = 1 × 10
-5
). No associations reached genome-wide
significance.

120
Supplemental Figure 4.16. Genome-Wide Association Manhattan Plot for Median
Fluorescence Intensity (MFI) of Antibodies to Viral Capsid Antigen p18 (VCA p18)
Among Participants Exceeding MFI Threshold for VCA p18 Positivity: Results from
University of Southern California Discovery Set

The genome-wide association study of VCA p18 MFI includes data from 131 University
of Southern California participants exceeding a 2526 VCA p18 MFI threshold and
approximately 7.8 million variants after imputation. The red line indicates genome-wide
significant associations (P = 5 × 10
-8
). The blue line indicates suggestive associations that
may warrant further investigation (P = 1 × 10
-5
). No associations reached genome-wide
significance.

121
Supplemental Figure 4.17. Genome-Wide Association Manhattan Plot for Median
Fluorescence Intensity (MFI) of Antibodies to Z-Epstein-Barr Virus Replication Activator
(ZEBRA) Among Participants Exceeding MFI Threshold for VCA ZEBRA Positivity:
Results from University of Southern California Discovery Set

The genome-wide association study of ZEBRA MFI includes data from 122 University of
Southern California participants exceeding a 74 ZEBRA MFI threshold and
approximately 7.8 million variants after imputation. The red line indicates genome-wide
significant associations (P = 5 × 10
-8
). The blue line indicates suggestive associations that
may warrant further investigation (P = 1 × 10
-5
). No associations reached genome-wide
significance.

122
Supplemental Figure 4.18. Quantile-Quantile Plots (-log10 Scale) for Median
Fluorescence Intensity of Antibodies to EBV Antigens: Results from University of
Southern California Discovery Set
Early antigen-diffuse (EA-D)
l = 0.99

EBV nuclear antigen 1 (EBNA-1)
l = 1.02

Viral capsid antigen p18 (VCA p18)
l = 1.01

Z-Epstein-Barr virus replication activator (ZEBRA)
l = 1.02

The genome-wide association studies of MFI levels of antibodies to EBV antigens
include data from University of Southern California participants exceeding corresponding
MFI thresholds and approximately 7.8 million variants after imputation. We did not
observe overdispersion of P values or inflation of the test statistic.

123

Chapter 5. DISCUSSION
SUMMARY OF FINDINGS
The projects included in this dissertation contribute a deeper understanding of the
epidemiology and immunopathology of Epstein-Barr virus (EBV) (2) through
investigation of the relationship between EBV, IM, and other immune-related conditions
including atopy, autoimmune disease, Hodgkin lymphoma, and non-Hodgkin lymphoma.
Each of the three projects included in this dissertation aimed to elucidate relationships
between the oncovirus and immune-related conditions.
In the first project, we used a cross-sectional approach to explore associations
between self-reported infectious mononucleosis (IM) and immune-related (autoimmune
and atopic) conditions among a population-based sample of twins. IM and atopy both
represent a deviation from the typical T-cell response to common immunological stimuli.
IM occurs in a subset of adolescents and young adults who experience delayed primary
EBV infection, and its symptoms appear to be associated with a disproportionate CD8+
and CD4+ T-cell response to primary EBV infection (20,25,31). Atopy is a genetic
predisposition to produce IgE antibodies in reaction to allergen exposure due to a T
helper type 2 response from CD4+ T helper cells (82). Matched analysis of twin pairs
discordant for IM did not reveal an association between IM and atopic or autoimmune
conditions. However, we demonstrated a strong positive relationship between self-
reported IM history and self-reported history of atopic conditions in multi-level models

124
accounting for individual- and twin pair-level effects. Among European American twin
pairs, the odds of IM were 1.71 times as high in twins reporting at least one atopic
condition compared to twins who did not (95% CI: 1.36, 2.16). Similarly, odds of IM
were 1.58 times as high in twins who reported a history of allergic rhinitis compared to
those who did not (95% CI: 1.24, 2.01). The coupling of findings from our multi-level
and matched-pair models indicate early childhood experiences may contribute to risk of
both IM and atopic conditions. These results are consistent with previous studies which
have shown that early childhood environment and parental attitudes toward childhood
microbe exposure may contribute to delayed EBV infection and incidence of atopy (99–
101).
In the second project, we examined potential interaction between self-reported IM
history and a candidate panel of 12 immune-related genetic variants on risk of non-
Hodgkin lymphoma and explored potential risk factors for IM among controls. Our
pooled case-control study suggested interaction between IM history and two variants on
T-cell lymphoma (TCL) risk: rs1143627 in interleukin-1B (IL1B) (Pinteraction = 0.04,
ORinteraction = 0.09, 95% confidence interval [CI] = 0.01, 0.87) and rs1800797 in
interleukin-6 (IL6) (Pinteraction = 0.03, ORinteraction = 0.08. 95% CI = 0.01, 0.80). Neither
interaction effect withstood adjustment for multiple comparisons. IL1B influences
inflammatory response through contributions to several lymphocyte activities including
growth and differentiation of B-cells (145), proliferation of T-helper Type 2 (Th2) clones
(146), and activation of Th17 cells (147). IL6 influences growth and differentiation of T-
cells, among many other immune functions (152,153). The minor alleles of the two IL1B

125
variants examined in our study are associated with lower expression of IL1B (150) and
may decrease T-cell activation in the setting of increased EBV load. This decrease in
activation may, in turn, attenuate the effect of IM on NHL. In our control population, we
confirmed a strong positive relationship between higher socio-economic status (SES) and
IM, which has been widely reported in other studies. These previous studies have
suggested high SES is a surrogate for a lower probability of EBV exposure in early life
and thus, a higher risk of developing IM (4,50,154,155). We also observed a strong
relationship between large sibship size and IM among controls born before 1960 (P =
0.001) but not in those born after 1960 (P = 0.85). Controls born after 1960 may have
been more likely to attend preschool, which would provide EBV exposure in early life
and simulate a large family. Alternatively, it may be that after 1960, overcrowding
decreased, and hygienic behaviors generally increased, mitigating the importance of
family size.
In the third project, we assessed the relationships between genetic variation and
EBV viral load among Hodgkin lymphoma patients through genome-wide association
studies (GWAS) and meta-analysis. EBV has been implicated in the pathogenesis of HL.
EBV proteins and viral DNA are present in Reed-Sternberg cells of 40% of HL cases in
the United States and Europe (59,60). Risk of EBV-positive HL was elevated in patients
who had been diagnosed with IM while risk of EBV-negative HL was not elevated after
IM diagnosis (61). We identified a missense variant at 3q29 in MUC4 associated with
EBV viral load (rs2246901; 95% confidence interval (CI): 473, 1572; P = 7.22 × 10
-9
).
Mucin 4 (MUC4), a transmembrane mucin protein encoded by the MUC4 gene, is

126
responsible for protecting and lubricating epithelial surfaces in the urogenital, digestive,
and respiratory tracts (194). Mucins impact inflammation and immune response via direct
interaction with leukocytes and other immune cells. Transmembrane mucins such as
MUC4 may hinder leukocyte motility and activation status and prevent the approach of
antigen-presenting cells (196). Variation in MUC4 may impact gene expression and the
availability of MUC4 on the epithelial surface in Waldeyer’s tonsillar ring: the primary
site of EBV infection and lytic reactivation. HL patients with higher MUC4 activation
may be unable to properly control EBV infection resulting in a higher EBV viral load. An
intronic variant at 3p26 in GRM7 (rs140444865; 95% CI: 0.20, 0.43; P = 5.47 × 10
-9
) and
an intergenic variant between INHA and STK11IP at 2q35 (rs115805790; 95% CI: 0.17,
0.41; P = 2.70 × 10
-8
) were associated with antibody levels to EBV nuclear antigen 1.
Metabotropic glutamate receptor 7 modulates neurotransmission, and polymorphisms in
GRM7 are related to neurodevelopmental disorders (199). The α subunit of inhibin
(INHA) is a member of the transforming growth factor beta (TGFβ) family, which
regulates angiogenesis. INHA has been implicated in tumor metastasis and
vascularization of ovarian cancers (200). Through its interaction with STK11,
serine/threonine kinase 11 interacting protein (STK11IP, alias LIP1) regulates STK11
function (cellular proliferation, signaling, and apoptosis (201,202).

127
FUTURE DIRECTIONS
Exploring the role of MUC4 variation in HL prognosis
Along with other mucins, MUC4, has been implicated in the prognosis of several
cancers of epithelial origin (195). Given our finding that variants in the MUC4 gene are
associated with EBV viral load, I would like to investigate the implications of these
mutations on response to various HL treatment modalities and failure-free survival. I am
also curious about whether the association between MUC4 variation and EBV viral load
extends to healthy European and non-European populations as well as those who have
other EBV-related conditions and neoplasms such as IM and specific non-Hodgkin
lymphoma subtypes. We can investigate these research questions through pooled analysis
and meta-analysis of data from InterLymph Consortium member sites.
Expanding the reach of infectious mononucleosis research to diverse populations
One limitation of this dissertation is the reliance of each project on data from
European/white participants. Although the findings offer insight into the role EBV plays
in immune conditions, we cannot be sure these results apply in other populations. I would
like to examine these relationships in ethnically diverse populations. Through my current
position as a data scientist for a large health insurance company, I have access to
healthcare claims data for tens of millions of customers. I would like to partner with my
committee chair, Dr. Wendy Cozen, and other PhD-level researchers at the insurance
company to profile IM incidence and severity in this patient population. I am particularly
interested in exploring differences in the profile of IM across sociodemographic factors
such as household income, geographic region, and race/ethnicity. The large population

128
will avail sufficient sample sizes to explore infectious mononucleosis in understudied
population strata.
In summary, this dissertation has contributed to an ever-growing body of evidence
linking poorly controlled Epstein-Barr virus infection to a host of immune-related
conditions. The timing of primary EBV infection as well as the immune response to
primary and persistent EBV infection provide insight about early childhood
immunological exposure and future immunological risk.

129
REFERENCES
1. Sprunt TP, Evans FA. Mononuclear leukocytosis in reaction to acute infections
(infectious mononucleosis). Bull Johns Hopkins Hosp. 1920;31:410–417.
2. Epstein M, Achong B, Barr Y. Virus Particles in Cultured Lymphoblasts from
Burkitt’s Lymphoma. Lancet. 1964;1(7335):702–703.
3. Luzuriaga K, Sullivan JL. Infectious mononucleosis. N Engl J Med.
2010;362(21):1993–2000.
4. Macsween KF, Johannessen I. Epstein-Barr Virus (EBV): Infectious
Mononucleosis and Other Non-malignant EBV-Associated Diseases. In: Kaslow
RA, Stanberry LR, Le Duc JW, eds. Viral Infections of Humans: Epidemiology
and Control. Boston, MA: Springer US; 2014:867–896.
5. Aronson M, Auwaerter P. Infectious Mononucleosis in adults and adolescents. In:
Hirsch MS, Kaplan SL, Bond S, eds. UpToDate. 2019:1–24.
6. Longnecker RM, Kieff E, Cohen JI. Epstein-Barr Virus. In: Knipe DM, Howley
PM, eds. Fields Virology. Philadelphia, PA: Lippincott Williams & Wilkins;
2013:1898–1959.
7. Parkin DM. The global health burden of infection-associated cancers in the year
2002. Int J Cancer. 2006;118(12):3030–3044.
8. Callan MFC. Epstein-Barr virus, arthritis, and the development of lymphoma in
arthritis patients. Curr Opin Rheumatol. 2004;16(4):399–405.
9. Ulff-Møller CJ, Nielsen NM, Rostgaard K, et al. Epstein-Barr virus-associated
infectious mononucleosis and risk of systemic lupus erythematosus. Rheumatology
(Oxford). 2010;49(9):1706–1712.
10. Thacker EL, Mirzaei F, Ascherio A. Infectious mononucleosis and risk for
multiple sclerosis: a meta-analysis. Ann Neurol. 2006;59(3):499–503.
11. Munger KL, Levin LI, O’Reilly EJ, et al. Anti-Epstein-Barr virus antibodies as
serological markers of multiple sclerosis: a prospective study among United States
military personnel. Mult Scler. 2011;17(10):1185–1193.
12. Lünemann JD, Tintoré M, Messmer B, et al. Elevated Epstein-Barr virus-encoded
nuclear antigen-1 immune responses predict conversion to multiple sclerosis. Ann
Neurol. 2010;67(2):159–169.
13. Liu F, Zhou ZH. Comparative virion structures of human herpesviruses. In: Arvin
A, Campadelli-Fiume G, Mocarski E, et al., eds. Human Herpesviruses: Biology,
Therapy, and Immunoprophylaxis. Cambridge: Cambridge University Press;

130
2007:1–25.
14. Pellett PE, Roizman B. Herpesviridae. In: Knipe DM, Howley PM, eds. Fields
Virology. Philadelphia, PA: Lippincott Williams & Wilkins; 2013:1802–1822.
15. Taylor GS, Long HM, Brooks JM, et al. The immunology of epstein-barr virus-
induced disease. Annu Rev Immunol. 2015;33:787–821.
16. Odumade OA, Hogquist KA, Balfour HH. Progress and problems in understanding
and managing primary epstein-barr virus infections. Clin Microbiol Rev.
2011;24(1):193–209.
17. Hislop AD, Taylor GS, Sauce D, et al. Cellular Responses to Viral Infection in
Humans: Lessons from Epstein-Barr Virus. Annu Rev Immunol. 2007;25(1):587–
617.
18. Macsween KF, Crawford DH. Epstein-Barr virus—recent advances. Lancet Infect
Dis. 2003;3(3):131–140.
19. Sullivan JL. Virology of Epstein-Barr virus - UpToDate. In: Hirsch MS, Edwards
MS, Bond S, eds. UpToDate. 2019:1–28.
20. Callan MFC, Steven N, Krausa P, et al. Large clonal expansions of CD8+ T cells
in acute infectious mononucleosis. Nat Med. 1996;2(8):906–911.
21. Abbott RJM, Quinn LL, Leese AM, et al. CD8 + T Cell Responses to Lytic EBV
Infection: Late Antigen Specificities as Subdominant Components of the Total
Response. J Immunol. 2013;191(11):5398–5409.
22. Callan MFC, Tan L, Annels N, et al. Direct visualization of antigen-specific CD8+
T cells during the primary immune response to Epstein-Barr virus in vivo. J Exp
Med. 1998;187(9):1395–1402.
23. Catalina MD, Sullivan JL, Bak KR, et al. Differential Evolution and Stability of
Epitope-Specific CD8 + T Cell Responses in EBV Infection. J Immunol.
2001;167(8):4450–4457.
24. Long HM, Chagoury OL, Leese AM, et al. Mhc II tetramers visualize human
CD4+ T cell responses to Epstein-Barr virus infection and demonstrate atypical
kinetics of the nuclear antigen EBNA1 response. J Exp Med. 2013;210(5):933–
949.
25. Long HM, Meckiff BJ, Taylor GS. The T-cell Response to Epstein-Barr Virus–
New Tricks From an Old Dog. Front Immunol. 2019;10:2193: 1–11.
26. Balfour HH, Holman CJ, Hokanson KM, et al. A prospective clinical study of
Epstein-Barr virus and host interactions during acute infectious mononucleosis. J
Infect Dis. 2005;192(9):1505–1512.

131
27. Hislop AD, Kuo M, Drake-Lee AB, et al. Tonsillar homing of Epstein-Barr virus-
specific CD8+ T cells and the virus-host balance. J Clin Invest. 2005;115(9):2546–
2555.
28. Biggar RJ, Henle G, Böcker J, et al. Primary epstein‐barr virus infections in
African infants. II. Clinical and serological observations during seroconversion. Int
J Cancer. 1978;22(3):244–250.
29. Fleisher G, Henle W, Henle G, et al. Primary Infection with Epstein-Barr Virus in
Infants in the United States: Clinical and Serologic Observations. J Infect Dis.
1979;139(5):553–558.
30. Silins SL, Sherritt MA, Silleri JM, et al. Asymptomatic primary Epstein-Barr virus
infection occurs in the absence of blood T-cell repertoire perturbations despite
high levels of systemic viral load. Blood. 2001;98(13):3739–3744.
31. Jayasooriya S, de Silva TI, Njie-jobe J, et al. Early Virological and Immunological
Events in Asymptomatic Epstein-Barr Virus Infection in African Children. PLOS
Pathog. 2015;11(3):e1004746.
32. Hwang AE, Hamilton AS, Cockburn MG, et al. Evidence of genetic susceptibility
to infectious mononucleosis: a twin study. Epidemiol Infect. 2012;140(11):2089–
2095.
33. Rostgaard K, Wohlfahrt J, Hjalgrim H. A genetic basis for infectious
mononucleosis: evidence from a family study of hospitalized cases in Denmark.
Clin Infect Dis. 2014;58:1684–1689.
34. McAulay KA, Higgins CD, Macsween KF, et al. HLA class I polymorphisms are
associated with development of infectious mononucleosis upon primary EBV
infection. J Clin Invest. 2007;117(10):3042–3048.
35. Ramagopalan S V, Meier UC, Conacher M, et al. Role of the HLA system in the
association between multiple sclerosis and infectious mononucleosis. Arch Neurol.
2011;68(4):469–72.
36. Tian C, Hromatka BS, Kiefer AK, et al. Genome-wide association and HLA region
fine-mapping studies identify susceptibility loci for multiple common infections.
Nat Commun. 2017;8(1):599: 1–13.
37. Helminen M, Lahdenpohja N, Hurme M. Polymorphism of the interleukin-10 gene
is associated with susceptibility to Epstein-Barr virus infection. J Infect Dis.
1999;180(Il):496–499.
38. Helminen ME, Kilpinen S, Virta M, et al. Susceptibility to primary Epstein-Barr
virus infection is associated with interleukin-10 gene promoter polymorphism. J
Infect Dis. 2001;184(6):777–80.

132
39. Balfour HH, Dunmire SK, Hogquist KA. Infectious mononucleosis. Clin Transl
Immunol. 2015;4(2):e33:1-7.
40. Macsween KF, Higgins CD, McAulay K a, et al. Infectious mononucleosis in
university students in the United Kingdom: evaluation of the clinical features and
consequences of the disease. Clin Infect Dis. 2010;50:699–706.
41. Krabbe S, Hesse J, Uldall P. Primary Epstein-Barr virus infection in early
childhood. Arch Dis Child. 1981;56(1):49–52.
42. Horwitz CA, Henle W, Henle G, et al. Clinical and laboratory evaluation of infants
and children with Epstein-Barr virus-induced infectious mononucleosis: report of
32 patients (aged 10-48 months). Blood. 1981;57(5):933–938.
43. Grotto I, Mimouni D, Huerta M, et al. Clinical and laboratory presentation of EBV
positive infectious mononucleosis in young adults. Epidemiol Infect.
2003;131(1):683–689.
44. Paul JR, Bunnell W. The presence of heterophile antibodies in infectious
mononucleosis. Am J Med Sci. 1932;183(1):90–103.
45. Biggar RJ, Gardner C, Lennette ET, et al. Malaria, Sex, and Place of Residence as
factors inn antibody response to Epstein-Barr Virus in Ghana, West Africa. Lancet.
1981;318(8238):115–118.
46. Hjalgrim H, Friborg J, Melbye M. The epidemiology of EBV and its association
with malignant disease. In: Arvin A, Campadelli-Fiume G, Mocarski E, et al., eds.
Human Herpesviruses: Biology, Therapy, and Immunoprophylaxis. Cambridge:
Cambridge University Press; 2007:1–53.
47. Pereira MS, Blake JM, Macrae AD. EB virus antibody at different ages. Br Med J.
1969;4(5682):526–527.
48. Hesse J, Ibsen KK, Krabbe S, et al. Prevalence of antibodies to Epstein-Barr virus
(EBV) in childhood and adolescence in Denmark. Scand J Infect Dis.
1983;15(4):335–338.
49. Lai PK, Mackay-Scollay EM, Alpers MP. Epidemiological studies of Epstein-Barr
herpesvirus infection in Western Australia. J Hyg (Lond). 1975;74(3):329–337.
50. Sumaya C V, Henle W, Henle G, et al. Seroepidemiologic study of Epstein-Barr
virus infections in a rural community. J Infect Dis. 1975;131(4):403–408.
51. Balfour HH, Odumade OA, Schmeling DO, et al. Behavioral, virologic, and
immunologic factors associated with acquisition and severity of primary Epstein-
Barr virus infection in university students. J Infect Dis. 2013;207(1):80–88.
52. Crawford DH, Macsween KF, Higgins CD, et al. A cohort study among university

133
students: identification of risk factors for Epstein-Barr virus seroconversion and
infectious mononucleosis. Clin Infect Dis. 2006;43:276–282.
53. Takeuchi K, Tanaka-Taya K, Kazuyama Y, et al. Prevalence of Epstein-Barr virus
in Japan: trends and future prediction. Pathol Int. 2006;56(3):112–116.
54. Morris MC, Edmunds WJ. The Changing Epidemiology of Infectious
Mononucleosis? J Infect. 2002;45:107–109.
55. Küppers R, Hansmann M-L. The Hodgkin and Reed/Sternberg cell. Int J Biochem
Cell Biol. 2005;37(3):511–517.
56. Jarrett RF. Viruses and Hodgkin’s lymphoma. Ann Oncol. 2002;13:23–29.
57. Levine PH, Ablashi D V, Berard CW, et al. Elevated antibody titers to epstein‐barr
virus in Hodgkin’s disease. Cancer. 1971;27(2):416–421.
58. Mueller N, Evans A, Harris NL, et al. Hodgkin’s disease and Epstein-Barr virus.
Altered antibody pattern before diagnosis. N Engl J Med. 1989;320(11):689–695.
59. Weiss LM, Movahed LA, Warnke RA, et al. Detection of Epstein–Barr Viral
Genomes in Reed–Sternberg Cells of Hodgkin’s Disease. N Engl J Med.
1989;320(8):502–506.
60. Pallesen G, Hamilton-Dutoit SJ, Rowe M, et al. Expression of Epstein-Barr virus
latent gene products in tumour cells of Hodgkin’s disease. Lancet (London,
England). 1991;337(8737):320–322.
61. Hjalgrim H, Askling J, Rostgaard K, et al. Characteristics of Hodgkin’s lymphoma
after infectious mononucleosis. N Engl J Med. 2003;349:1324–1332.
62. Glaser SL, Lin RJ, Stewart SL, et al. Epstein-Barr virus-associated Hodgkin’s
disease: Epidemiologic characteristics in international data. Int J Cancer.
1997;70(4):375–382.
63. Urayama KY, Jarrett RF, Hjalgrim H, et al. Genome-wide association study of
classical hodgkin lymphoma and epstein-barr virus status-defined subgroups. J
Natl Cancer Inst. 2012;104(3):240–253.
64. Cozen W, Timofeeva MN, Li D, et al. A meta-analysis of Hodgkin lymphoma
reveals 19p13.3 TCF3 as a novel susceptibility locus. Nat Commun.
2014;5(1):3856–3865.
65. Gandhi MK, Lambley E, Burrows J, et al. Plasma Epstein-Barr virus (EBV) DNA
is a biomarker for EBV-positive Hodgkin’s lymphoma. Clin Cancer Res.
2006;12(2):460–464.
66. Spacek M, Hubacek P, Markova J, et al. Plasma EBV-DNA monitoring in Epstein-

134
Barr virus-positive Hodgkin lymphoma patients. APMIS. 2011;119(1):10–16.
67. Kanakry JA, Li H, Gellert LL, et al. Plasma Epstein-Barr virus DNA predicts
outcome in advanced Hodgkin lymphoma: correlative analysis from a large North
American cooperative group trial. Blood. 2013;121(18):3547–3553.
68. Morton LM, Sampson JN, Cerhan JR, et al. Rationale and Design of the
International Lymphoma Epidemiology Consortium (InterLymph) Non-Hodgkin
Lymphoma Subtypes Project. J Natl Cancer Inst - Monogr. 2014;2014(48):1–14.
69. Moormann AM, Bailey JA. Malaria — how this parasitic infection aids and abets
EBV-associated Burkitt lymphomagenesis. Curr Opin Virol. 2016;78–84.
70. Farrell PJ. Epstein–Barr Virus and Cancer. Annu Rev Pathol Mech Dis.
2019;14(1):29–53.
71. Jones JF, Shurin S, Abramowsky C, et al. T-Cell Lymphomas Containing Epstein–
Barr Viral DNA in Patients with Chronic Epstein–Barr Virus Infections. N Engl J
Med. 1988;318(12):733–741.
72. Chen CL, Sadler RH, Walling DM, et al. Epstein-Barr virus (EBV) gene
expression in EBV-positive peripheral T-cell lymphomas. J Virol.
1993;67(10):6303–6308.
73. Park S, Lee J, Young HK, et al. The impact of Epstein-Barr virus status on clinical
outcome in diffuse large B-cell lymphoma. Blood. 2007;110(3):972–978.
74. Piriou ER, van Dort K, Nanlohy NM, et al. Novel method for detection of virus-
specific CD4+ T cells indicates a decreased EBV-specific CD4+ T cell response in
untreated HIV-infected subjects. Eur J Immunol. 2005;35(3):796–805.
75. International Collaboration on HIV and Cancer. Highly Active Antiretroviral
Therapy and Incidence of Cancer in Human Immunodeficiency Virus-Infected
Adults. J Natl Cancer Inst. 2000;92(22):1823–1830.
76. Faulkner GC, Burrows SR, Khanna R, et al. X-Linked agammaglobulinemia
patients are not infected with Epstein-Barr virus: implications for the biology of
the virus. J Virol. 1999;73(2):1555–1564.
77. Rose NR. Overview of autoimmunity. In: Schur PH, Romain PL, eds. UpToDate.
2019:1–19.
78. Davidson A, Diamond B. Advances in Immunology: Autoimmune diseases. N
Engl J Med. 2001;345(5):340–350.
79. Ascherio A, Munger KL. EBV and Autoimmunity. In: Münz C, ed. Epstein Barr
Virus Volume 1. Current Topics in Microbiology and Immunology. Springer,
Cham; 2015:365–385.

135
80. Nielsen TR, Rostgaard K, Nielsen NM, et al. Multiple Sclerosis After Infectious
Mononucleosis. Arch Neurol. 2007;64(1):72–75.
81. Tosato G, Steinberg AD, Blaese RM. Defective EBV-specific suppressor T-cell
function in rheumatoid arthritis. N Engl J Med. 1981;305(21):1238–1243.
82. Cockcroft DW. Allergens. In: Barnes PJ, Drazen JM, Rennard SI, et al., eds.
Asthma and COPD. Academic Press; 2009:443–455.
83. Mosmann TR, Coffman RL. TH1 and TH2 Cells: Different Patterns of
Lymphokine Secretion Lead to Different Functional Properties. Annu Rev
Immunol. 1989;7(1):145–173.
84. Oettgen HC. Fifty years later: Emerging functions of IgE antibodies in host
defense, immune regulation, and allergic diseases. J Allergy Clin Immunol.
2016;137(6):1631–1645.
85. Heimall J. The adaptive cellular immune response: T cells and cytokines -
UpToDate. In: Notarangelo LD, TePas E, eds. UpToDate. 2019:1–30.
86. Stokes J, Casale TB. The relationship between IgE and allergic disease -
UpToDate. In: Bochner BS, Feldweg AM, eds. UpToDate. 2019:1–21.
87. Klink M, Cline MG, Halonen M, et al. Problems in defining normal limits for
serum IgE. J Allergy Clin Immunol. 1990;85(2):440–444.
88. Nilsson C, Linde A, Montgomery SM, et al. Does early EBV infection protect
against IgE sensitization? J Allergy Clin Immunol. 2005;116(2):438–444.
89. Saghafian-Hedengren S, Sverremark-Ekström E, Linde A, et al. Early-life EBV
infection protects against persistent IgE sensitization. J Allergy Clin Immunol.
2010;125(2):433–438.
90. Veiga R V, Cunha SS, Dattoli VC, et al. Chronic virus infections supress atopy but
not asthma in a set of children from a large latin american city: a cross-section
study. BMC Pulm Med. 2011;11(1):24.
91. Illi S, Mutius E Von, Lau S, et al. Early childhood infectious diseases and the
development of asthma up to school age : a birth cohort study. BMJ.
2001;322(7283):390–395.
92. Balfour HH, Sifakis F, Sliman JA, et al. Age-Specific prevalence of epstein-barr
virus infection among individuals aged 6-19 years in the United States and factors
affecting its acquisition. J Infect Dis. 2013;208(8):1286–1293.
93. Evans AS. Infectious mononucleosis and related syndromes. Am J Med Sci.
1978;276(3):325–39.

136
94. Cockburn MG, Hamilton AS, Zadnick J, et al. Development and
representativeness of a large population-based cohort of native Californian twins.
Twin Res. 2001;4(4):242–250.
95. R Core Team. R: A language and environment for statistical computing. 2018;
96. Bates D, Mächler M, Bolker BM, et al. Fitting linear mixed-effects models using
lme4. J Stat Softw. 2015;67(1):1–48.
97. Therneau TM. A Package for Survival Analysis in S. Version 2.38. 2015;
98. Venables WN, Ripley BD. Modern Applied Statistics with S Fourth edition by.
Fourth. New York: Springer; 2002.
99. Gares V, Panico L, Castagne R, et al. The role of the early social environment on
Epstein Barr virus infection: A prospective observational design using the
Millennium Cohort Study. Epidemiol Infect. 2017;145(16):3405–3412.
100. Thompson JC, Dolen WK. Pacifier cleaning practices and risk of allergy
development. Pediatrics. 2014;131(6):e1829–e1837.
101. Hesselmar B, Hicke-Roberts A, Wennergren G. Allergy in children in hand versus
machine dishwashing. Pediatrics. 2015;135(3):e590-597.
102. Ober C, Yao TC. The genetics of asthma and allergic disease: A 21st century
perspective. Immunol Rev. 2011;242(1):10–30.
103. Yang J, Qiao HL, Zhang YW, et al. HLA-DRB genotype and specific IgE
responses in patients with allergies to penicillins. Chin Med J (Engl).
2006;119(6):458–466.
104. Demenais F, Margaritte-Jeannin P, Barnes KC, et al. Multiancestry association
study identifies new asthma risk loci that colocalize with immune-cell enhancer
marks. Nat Genet. 2018;50(1):42–53.
105. Kreiner E, Waage J, Standl M, et al. Shared genetic variants suggest common
pathways in allergy and autoimmune diseases. J Allergy Clin Immunol.
2017;140(3):771–781.
106. Burbank AJ, Sood AK, Kesic MJ, et al. Environmental determinants of allergy and
asthma in early life. J Allergy Clin Immunol. 2017;140(1):1–12.
107. Grulich AE, Vajdic CM, Cozen W. Altered Immunity as a Risk Factor for Non-
Hodgkin Lymphoma. Cancer Epidemiol Prev Biomarkers. 2007;16(3):405–409.
108. Morton LM, Wang SS, Cozen W, et al. Etiologic heterogeneity among non-
Hodgkin lymphoma subtypes. Blood. 2008;112(13):5150–5160.
109. Smedby KE, Vajdic CM, Falster M, et al. Autoimmune disorders and risk of non-

137
Hodgkin lymphoma subtypes: A pooled analysis within the InterLymph
Consortium. Blood. 2008;111:4029–4038.
110. Cozen W, Cerhan JR, Martinez-Maza O, et al. The effect of atopy, childhood
crowding, and other immune-related factors on non-Hodgkin lymphoma risk.
Cancer Causes Control. 2007;18:821–831.
111. Goldin LR, Björkholm M, Kristinsson SY, et al. Highly increased familial risks for
specific lymphoma subtypes. Br J Haematol. 2009;146(1):91–94.
112. Wang SS, Flowers CR, Kadin ME, et al. Medical history, lifestyle, family history,
and occupational risk factors for peripheral T-cell lymphomas: The interlymph
non-hodgkin lymphoma subtypes project. J Natl Cancer Inst - Monogr.
2014;2014(48):66–75.
113. Hosgood HD, Purdue MP, Wang SS, et al. A pooled analysis of three studies
evaluating genetic variation in innate immunity genes and non-Hodgkin lymphoma
risk. Br J Haematol. 2011;152(6):721–726.
114. Skibola CF, Akers NK, Conde L, et al. Multi-locus HLA class I and II allele and
haplotype associations with follicular lymphoma. Tissue Antigens.
2012;79(4):279–286.
115. Wang SS, Abdou AM, Morton LM, et al. Human leukocyte antigen class I and II
alleles in non-Hodgkin lymphoma etiology. Blood. 2010;115:4820–4823.
116. Thorley-Lawson DA, Gross A. Persistence of the Epstein–Barr Virus and the
Origins of Associated Lymphomas. N Engl J Med. 2004;350(13):1328–1337.
117. Ferri C, Caracciolo F, Zignego AL, et al. Hepatitis C virus infection in patients
with non-Hodgkin’s lymphoma. Br J Haematol. 1994;88(2):392–394.
118. Parsonnet J, Hansen S, Rodriguez L, et al. Helicobacter pylori infection and gastric
lymphoma. N Engl J Med. 1994;330(18):1267–1271.
119. Engels EA. Infectious agents as causes of non-Hodgkin lymphoma. Cancer
Epidemiol Biomarkers Prev. 2007;16(March):401–404.
120. Becker N, Falster MO, Vajdic CM, et al. Self-reported history of infections and the
risk of non-Hodgkin lymphoma: An InterLymph pooled analysis. Int J Cancer.
2012;131(May 2011):2342–2348.
121. Kane E, Skibola CF, Bracci PM, et al. Non-Hodgkin Lymphoma, Body Mass
Index, and Cytokine Polymorphisms: A Pooled Analysis from the InterLymph
Consortium. Cancer Epidemiol Biomarkers Prev. 2015;24(7):1061–1070.
122. Skibola CF, Bracci PM, Nieters A, et al. Tumor necrosis factor (TNF) and
lymphotoxin-alpha (LTA) polymorphisms and risk of non-Hodgkin lymphoma in

138
the InterLymph Consortium. Am J Epidemiol. 2010;171:267–276.
123. Rothman N, Skibola CF, Wang SS, et al. Genetic variation in TNF and IL10 and
risk of non-Hodgkin lymphoma: a report from the InterLymph Consortium. Lancet
Oncol. 2006;7(1):27–38.
124. Cerhan JR, Ansell SM, Fredericksen ZS, et al. Genetic variation in 1253 immune
and inflammation genes and risk of non-Hodgkin lymphoma. Blood.
2007;110(13):4455–4463.
125. Cerhan JR, Fredericksen ZS, Novak AJ, et al. A two-stage evaluation of genetic
variation in immune and inflammation genes with risk of non-Hodgkin lymphoma
identifies new susceptibility locus in 6p21.3 region. Cancer Epidemiol Biomarkers
Prev. 2012;21(10):1799–1806.
126. Besson H, Brennan P, Becker N, et al. Tobacco smoking, alcohol drinking and
non-Hodgkin’s lymphoma: A European multicenter case-control study
(Epilymph). Int J Cancer. 2006;119(4):901–908.
127. Becker N, Fortuny J, Alvaro T, et al. Medical history and risk of lymphoma:
Results of a European case-control study (EPILYMPH). J Cancer Res Clin Oncol.
2009;135:1099–1107.
128. Cerhan JR, Fredericksen ZS, Wang AH, et al. Design and validity of a clinic-based
case-control study on the molecular epidemiology of lymphoma. Int J Mol
Epidemiol Genet. 2011;2(2):95–113.
129. Chatterjee N, Hartge P, Cerhan JR, et al. Risk of non-Hodgkin’s lymphoma and
family history of lymphatic, hematologic, and other cancers. Cancer Epidemiol
Biomarkers Prev. 2004;13(9):1415–1421.
130. Hughes AM, Armstrong BK, Vajdic CM, et al. Sun exposure may protect against
non-Hodgkin lymphoma: a case-control study. Int J Cancer. 2004;112(5):865–71.
131. Smedby KE, Hjalgrim H, Melbye M, et al. Ultraviolet radiation exposure and risk
of malignant lymphomas. J Natl Cancer Inst. 2005;97(3):199–209.
132. Holly EA, Lele C, Bracci PM, et al. Case-control study of non-Hodgkin’s
lymphoma among women and heterosexual men in the San Francisco Bay Area,
California. Am J Epidemiol. 1999;150(4):375–89.
133. Holly EA, Bracci PM. Population-based study of non-Hodgkin lymphoma,
histology, and medical history among human immunodeficiency virus-negative
participants in San Francisco. Am J Epidemiol. 2003;158(4):316–27.
134. Morton LM, Holford TR, Leaderer B, et al. Alcohol use and risk of non-Hodgkin’s
lymphoma among Connecticut women (United States). Cancer Causes Control.
2003;14(7):687–694.

139
135. Conde L, Halperin E, Akers NK, et al. Genome-wide association study of
follicular lymphoma identifies a risk locus at 6p21.32. Nat Genet. 2010;42:661–
664.
136. Spinelli JJ, Ng CH, Weber J-P, et al. Organochlorines and risk of non-Hodgkin
lymphoma. Int J Cancer. 2007;121(12):2767–75.
137. Smedby KE, Foo JN, Skibola CF, et al. GWAS of follicular lymphoma reveals
allelic heterogeneity at 6p21.32 and suggests shared genetic susceptibility with
diffuse large B-cell lymphoma. PLoS Genet. 2011;7(4):e1001378: 1-11.
138. Morton LM, Turner JJ, Cerhan JR, et al. Proposed classification of lymphoid
neoplasms for epidemiologic research from the Pathology Working Group of the
International Lymphoma Epidemiology Consortium (InterLymph). Blood.
2007;110(2):695–708.
139. Turner JJ, Morton LM, Linet MS, et al. InterLymph hierarchical classification of
lymphoid neoplasms for epidemiologic research based on the WHO classification
(2008): Update and future directions. Blood. 2010;116(20):e90-398.
140. Swerdlow SH, Campo E, Harris NL, et al. WHO Classification of Tumours of
Haematopoietic and Lymphoid Tissues. 4th ed. International Agency for Research
on Cancer; 2008.
141. Johnson AD, Handsaker RE, Pulit SL, et al. SNAP: A web-based tool for
identification and annotation of proxy SNPs using HapMap. Bioinformatics.
2008;24(24):2938–2939.
142. Mukherjee B, Ahn J, Gruber SB, et al. Tests for gene-environment interaction
from case-control data: a novel study of type I error, power and designs. Genet
Epidemiol. 2008;32(7):615–626.
143. Kamerman SB. A global history of early childhood education and care. 2006 1–92
p.
144. Conneely KN, Boehnke M. So Many Correlated Tests, So Little Time! Rapid
Adjustment of P Values for Multiple Correlated Tests. Am J Hum Genet.
2007;81(6):1158–1168.
145. Pike BL, Nossal GJ. Interleukin 1 can act as a B-cell growth and differentiation
factor. Proc Natl Acad Sci U S A. 1985;82(23):8153–8157.
146. Lichtman AH, Chin J, Schmidt JA, et al. Role of interleukin 1 in the activation of
T lymphocytes. Proc Natl Acad Sci U S A. 1988;85(24):9699–9703.
147. Schett G, Dayer J-M, Manger B. Interleukin-1 function and role in rheumatic
disease. Nat Rev Rheumatol. 2016;12(1):14–24.

140
148. Nambu A, Nakae S, Iwakura Y. IL-1β, but not IL-1α, is required for antigen-
specific T cell activation and the induction of local inflammation in the delayed-
type hypersensitivity responses. Int Immunol. 2006;18(5):701–712.
149. Hackett RJ, Davis LS, Lipsky PE. Comparative effects of tumor necrosis factor-
alpha and IL-1 beta on mitogen-induced T cell activation. J Immunol.
1988;140(8):2639–2644.
150. Hirbod-Mobarakeh A, Amirzargar AA, Nikbin B, et al. Immunogenetics of
Cancer. In: Cancer Immunology. Berlin, Heidelberg: Springer Berlin Heidelberg;
2015:295–341.
151. Kasztelewicz B, Jankowska I, Pawłowska J, et al. The impact of cytokine gene
polymorphisms on Epstein-Barr virus infection outcome in pediatric liver
transplant recipients. J Clin Virol. 2012;55(3):226–232.
152. Terry CF, Loukaci V, Green FR. Cooperative influence of genetic polymorphisms
on interleukin 6 transcriptional regulation. J Biol Chem. 2000;275(24):18138–
18144.
153. Burger R. Impact of Interleukin-6 in hematological malignancies. Transfus Med
Hemotherapy. 2013;40(5):336–343.
154. Nye FJ. Social class and infectious mononucleosis. J Hyg (Lond). 1973;71(1):145–
149.
155. Lehane DE. A Seroepidemiologic Study of Infectious Mononucleosis. JAMA.
1970;212(13):2240.
156. Rostgaard K, Nielsen TR, Wohlfahrt J, et al. Sibship structure and risk of
infectious mononucleosis: a population-based cohort study. Int J Epidemiol.
2014;43(5):1607–1614.
157. Liu Z, Fang F, Chang ET, et al. Sibship size, birth order and risk of
nasopharyngeal carcinoma and infectious mononucleosis: a nationwide study in
Sweden. Int J Epidemiol. 2016;45(3):825–834.
158. Ramagopalan S V, Giovannoni G, Yeates DG, et al. Sex ratio of infectious
mononucleosis and possible relevance to multiple sclerosis. Mult Scler J.
2013;19(3):359–361.
159. Petersen I, Thomas JM, Hamilton WT, et al. Risk and predictors of fatigue after
infectious mononucleosis in a large primary-care cohort. QJM - Mon J Assoc
Physicians. 2006;99(December 2005):49–55.
160. Buchwald DS, Rea TD, Katon WJ, et al. Acute infectious mononucleosis:
characteristics of patients who report failure to recover. Am J Med.
2000;109(00):531–537.

141
161. Beagley KW, Gockel CM. Regulation of innate and adaptive immunity by the
female sex hormones oestradiol and progesterone. FEMS Immunol Med Microbiol.
2003;38(1):13–22.
162. Fink AL, Klein SL. Sex and Gender Impact Immune Responses to Vaccines
Among the Elderly. Physiology. 2015;30(6):408–416.
163. Abramson JS, Feldman T, Kroll-Desrosiers AR, et al. Peripheral T-cell
lymphomas in a large US multicenter cohort: Prognostication in the modern era
including impact of frontline therapy. Ann Oncol. 2014;25(11):2211–2217.
164. Coiffier B, Brousse N, Peuchmaur M, et al. Original article: Peripheral T-cell
lymphomas have a worse prognosis than B-cell lymphomas: A prospective study
of 361 immunophenotyped patients treated with the LNH-84 regimen. Ann Oncol.
1990;1(1):45–50.
165. Weisenburger DD, Savage KJ, Harris NL, et al. Peripheral T-cell lymphoma, not
otherwise specified: A report of 340 cases from the International Peripheral T-cell
Lymphoma Project. Blood. 2011;117(12):3402–3408.
166. Talpur R, Singh L, Daulat S, et al. Long-term outcomes of 1,263 patients with
mycosis fungoides and sézary syndrome from 1982 to 2009. Clin Cancer Res.
2012;18(18):5051–5060.
167. Houldcroft CJ, Kellam P. Host genetics of Epstein-Barr virus infection, latency
and disease. Rev Med Virol. 2015;25(2):71–84.
168. Enciso-Mora V, Broderick P, Ma Y, et al. A genome-wide association study of
Hodgkin’s lymphoma identifies new susceptibility loci at 2p16.1 (REL), 8q24.21
and 10p14 (GATA3). Nat Genet. 2010;42(12):1126–1130.
169. Best T, Li D, Skol AD, et al. Variants at 6q21 implicate PRDM1 in the etiology of
therapy-induced second malignancies after Hodgkin’s lymphoma. Nat Med.
2011;17(8):941–943.
170. Cozen W, Li D, Best T, et al. A genome-wide meta-analysis of nodular sclerosing
Hodgkin lymphoma identifies risk loci at 6p21.32. Blood. 2012;119(2):469–475.
171. Khankhanian P, Cozen W, Himmelstein DS, et al. Meta-analysis of genome-wide
association studies reveals genetic overlap between Hodgkin lymphoma and
multiple sclerosis. Int J Epidemiol. 2016;45(3):728–40.
172. Hwang AE, Marshall V, Conti D V., et al. Epstein–Barr virus load is higher in
long-term Hodgkin lymphoma survivors compared to their unaffected twins and
unrelated controls. Br J Haematol. 2019;185(2):377–380.
173. Besson C, Amiel C, Le‐Pendeven C, et al. Strong Correlations of Anti–Viral
Capsid Antigen Antibody Levels in First‐Degree Relatives from Families with

142
Epstein‐Barr Virus–Related Lymphomas. J Infect Dis. 2009;199(8):1121–1127.
174. Pedergnana V, Syx L, Cobat A, et al. Combined linkage and association studies
show that HLA class II variants control levels of antibodies against Epstein-Barr
virus antigens. PLoS One. 2014;
175. Besson C, Amiel C, Le-Pendeven C, et al. Positive correlation between Epstein-
Barr virus viral load and anti-viral capsid immunoglobulin G titers determined for
Hodgkin’s lymphoma patients and their relatives. J Clin Microbiol.
2006;44(1):47–50.
176. Dehee A, Asselot C, Piolot T, et al. Quantification of Epstein-Barr virus load in
peripheral blood of human immunodeficiency virus-infected patients using real-
time PCR. J Med Virol. 2001;65(3):543–552.
177. Los Angeles - SEER Registries. Natl Cancer Inst Surveillance, Epidemiol End
Results Progr. (https://seer.cancer.gov/registries/los_angeles.html). (Accessed
December 2, 2019)
178. San Francisco-Oakland - SEER Registries. Natl Cancer Inst Surveillance,
Epidemiol End Results Progr. (https://seer.cancer.gov/registries/sfo.html).
(Accessed December 2, 2019)
179. San Jose - Monterey - SEER Registries. Natl Cancer Inst Surveillance, Epidemiol
End Results Progr. (https://seer.cancer.gov/registries/sjm.html). (Accessed
December 2, 2019)
180. Mack T, Deapen D, Hamilton AS. Representativeness of a roster of volunteer
North American twins with chronic disease. Twin Res. 2000;3(1):33–42.
181. Das S, Forer L, Schönherr S, et al. Next-generation genotype imputation service
and methods. Nat Genet. 2016;48(10):1284–1287.
182. Chin Yuan C, Miley W, Waters D. A quantification of human cells using an ERV-
3 real time PCR assay. J Virol Methods. 2001;91(2):109–117.
183. Brenner N, Mentzer AJ, Butt J, et al. Validation of Multiplex Serology detecting
human herpesviruses 1-5. PLoS One. 2018;13(12):e0209379: 1-20.
184. Pruim RJ, Welch RP, Sanna S, et al. LocusZoom: Regional visualization of
genome-wide association scan results. In: Bioinformatics. Narnia; 2011:2336–
2337.
185. Stram DO. Correcting for Hidden Population Structure in Single Marker
Association Testing and Estimation. In: Gail M, Krickeberg K, Samet J, et al., eds.
Design, Analysis, and Interpretation of Genome-Wide Association Scans. New
York, NY: Springer; 2014:135–181.

143
186. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic
variants from high-throughput sequencing data. Nucleic Acids Res.
2010;38(16):e164–e164.
187. Schneider V, Church D. Genome Reference Consortium. In: Beck J, Benson D,
Coleman J, et al., eds. The NCBI Handbook [Internet]. National Center for
Biotechnology Information (US); 2013
188. O’Leary NA, Wright MW, Brister JR, et al. Reference sequence (RefSeq) database
at NCBI: current status, taxonomic expansion, and functional annotation. Nucleic
Acids Res. 2016;44(D1):D733–D745.
189. Chang CC, Chow CC, Tellier LC, et al. Second-generation PLINK: Rising to the
challenge of larger and richer datasets. Gigascience. 2015;4(7):1–16.
190. Turner SD. qqman: an R package for visualizing GWAS results using Q-Q and
manhattan plots. J Open Source Softw. 2018;3(25):731: 1–2.
191. Turner S. qqman: Q-Q and Manhattan Plots for GWAS Data. 2017;
192. Zhao JH. gap: Genetic Analysis Package. J Stat Softw. 2007;23(8):1–18.
193. Zhao JH. gap: Genetic Analysis Package. 2019;
194. Singh AP, Chaturvedi P, Batra SK. Emerging Roles of MUC4 in Cancer: A Novel
Target for Diagnosis and Therapy. Cancer Res. 2007;67(2):433–436.
195. Jonckheere N, Van Seuningen I. Integrative analysis of the cancer genome atlas
and cancer cell lines encyclopedia large-scale genomic databases:
MUC4/MUC16/MUC20 signature is associated with poor survival in human
carcinomas. J Transl Med. 2018;16(1):259: 1–22.
196. Hollingsworth MA, Swanson BJ. Mucins in cancer: protection and control of the
cell surface. Nat Rev Cancer. 2004;4(1):45–60.
197. Friedlaender A, Addeo A, Banna G. New emerging targets in cancer
immunotherapy: the role of TIM3. ESMO Open. 2019;4(Suppl 3):e000497.
198. Kondo S, Yoshizaki T, Wakisaka N, et al. MUC1 induced by Epstein-Barr virus
latent membrane protein 1 causes dissociation of the cell-matrix interaction and
cellular invasiveness via STAT signaling. J Virol. 2007;81(4):1554–62.
199. Fisher NM, Seto M, Lindsley CW, et al. Metabotropic Glutamate Receptor 7: A
New Therapeutic Target in Neurodevelopmental Disorders. Front Mol Neurosci.
2018;11:387.
200. Singh P, Jenkins LM, Horst B, et al. Inhibin Is a Novel Paracrine Factor for Tumor
Angiogenesis and Metastasis. Cancer Res. 2018;78(11):2978–2989.

144
201. Smith DP, Rayter SI, Niederlander C, et al. LIP1, a cytoplasmic protein
functionally linked to the Peutz-Jeghers syndrome kinase LKB1. Hum Mol Genet.
2001;10(25):2869–2877.
202. Beggs AD, Latchford AR, Vasen HFA, et al. Peutz-Jeghers syndrome: a
systematic review and recommendations for management. Gut. 2010;59(7):975–
86.

Abstract (if available)

Abstract Epstein-Barr virus (EBV) is a ubiquitous human herpes virus which is estimated to infect over 90% of adults worldwide. Infectious mononucleosis (IM), a clinical syndrome typified by fever, pharyngitis, and lymphadenopathy, is a response to primary EBV infection. IM is most common among individuals whose primary EBV infection occurs during adolescence or young adulthood. In addition to serious sequelae related to IM, EBV is associated with several malignancies. EBV DNA is present in 95–100% of Burkitt lymphomas in endemic regions, 100% of anaplastic nasopharyngeal carcinomas, 60–90% of mixed cellularity and lymphocyte-depleted Hodgkin lymphoma subtypes, 100% of lymphoepitheliomas, and 25–100% of various non-Hodgkin lymphoma subtypes. By conservative estimates, at least 113,000 incident cases of cancer each year, roughly 1% of worldwide cancer incidence, are attributable to EBV infection and could be prevented if EBV was eliminated or successfully controlled by the immune system. Despite the established associations with immunopathology and malignancy, questions remain about the role of environmental, genetic, and immunological factors in determining the severity of primary EBV infection, the implications of symptomatic infection for future disease risk, and the role of genetic variation in symptomatic primary and reactivated (replicating) EBV infection. This dissertation includes three projects aimed at better framing the relationship between EBV, IM, and other immune-related conditions including autoimmune disease, atopy, Hodgkin lymphoma, and non-Hodgkin lymphoma.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Common immune-related factors and risk of non-Hodgkin lymphomy

PDF

Early childhood health experience & adult phenotype in twins

PDF

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

PDF

Genetic risk factors in multiple myeloma

PDF

Genetic studies of cancer in populations of African ancestry and Latinos

PDF

Genetic epidemiological approaches in the study of risk factors for hematologic malignancies

PDF

Lifestyle-related exposures and diseases in twins

PDF

Prostate cancer: genetic susceptibility and lifestyle risk factors

PDF

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

PDF

Dietary carcinogens and genetic variation in their metabolism: epidemiological studies on the risk of selected cancers

PDF

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

PDF

Genetic studies of inflammation and cardiovascular disease

PDF

Personal hair dye use and risk of B-cell non-Hodgkin’s lymphomas among adult women in Los Angeles County

PDF

Environmental risk factors of Multiple Sclerosis: a twin study

PDF

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

PDF

Antibiotic resistance in a large ICU cohort: 1995-2002

PDF

Dietary fats, atherosclerosis, and cardiovascular disease

PDF

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

PDF

Investigating the modulatory mechanisms of ILC2s as a therapeutic strategy in the context of health and disease

PDF

Comparison of participant and study partner predictions of cognitive impairment in the Alzheimer's disease neuroimaging initiative 3 study

Asset Metadata

Creator Wadé, Niquelle Brown (author)

Core Title Epidemiological studies of Epstein-Barr virus, lymphoma, and immune conditions

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Publication Date 03/23/2020

Defense Date 12/17/2019

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

Tag allergic rhinitis,atopy,autoimmune disease,Epstein-Barr virus,GWAS,Hodgkin lymphoma,Infectious Mononucleosis,non-Hodgkin lymphoma,OAI-PMH Harvest

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cozen, Wendy (committee chair), Conti, David (committee member), Jarrett, Ruth (committee member), Mack, Thomas (committee member)

Creator Email niquelle.brown@gmail.com,niquelle.bw@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-276941

Unique identifier UC11673147

Identifier etd-WadeNiquel-8227.pdf (filename),usctheses-c89-276941 (legacy record id)

Legacy Identifier etd-WadeNiquel-8227.pdf

Dmrecord 276941

Document Type Dissertation

Rights Wadé, Niquelle Brown; Wade, Niquelle Brown

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