University of Southern California Dissertations and Theses

The role of inflammation in non-Hodgkin lymphoma etiology

(USC Thesis Other)

The role of inflammation in non-Hodgkin lymphoma etiology

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

The Role of Inflammation in Non-Hodgkin Lymphoma Etiology

by
Charlie Zhong, MS MPH

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)

December 2019
ii

Table of Contents
List of Tables ................................................................................................................................................ iv
List of Figures ............................................................................................................................................... vi
List of Abbreviations ................................................................................................................................... vii
Chapter 1. Background ................................................................................................................................. 1
Chapter 2. The Human Leukocyte Antigen and Non-Hodgkin Lymphoma Risk Among Transplant
Indicated Patients Seen at The City of Hope National Medical Center ........................................................ 4
2.1. Introduction ....................................................................................................................................... 4
2.1.1. Diffuse Large B-cell Lymphoma ................................................................................................... 6
2.1.2. Follicular Lymphoma ................................................................................................................. 10
2.1.3. Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma .............................................. 15
2.1.4. Marginal Zone Lymphoma ........................................................................................................ 20
2.2. Methods ........................................................................................................................................... 21
2.2.1. Study Population ....................................................................................................................... 21
2.2.2. NHL definition ........................................................................................................................... 21
2.2.3. HLA Typing and Imputation ...................................................................................................... 22
2.2.4. Analytic Population ................................................................................................................... 23
2.2.5. Statistical Analysis ..................................................................................................................... 24
2.3. Results .............................................................................................................................................. 24
2.3.1. Diffuse Large B-cell Lymphoma ................................................................................................. 25
2.3.2. Follicular Lymphoma ................................................................................................................. 26
2.3.3. Marginal Zone Lymphoma ........................................................................................................ 26
2.3.4. HLA Class I and II homozygosity ................................................................................................ 27
2.4. Discussion ......................................................................................................................................... 27
Chapter 3. Artificial Light at Night at Non-Hodgkin Lymphoma Risk in the California Teachers Study
Cohort ......................................................................................................................................................... 52
3.1. Introduction ..................................................................................................................................... 52
3.2. Methods ........................................................................................................................................... 55
3.2.1. Study Population ....................................................................................................................... 55
3.2.2. Exposure Assessment ................................................................................................................ 56
3.2.3. Outcome Assessment ............................................................................................................... 57
3.2.4. Covariates ................................................................................................................................. 57
iii

3.2.5. Statistical Methods ................................................................................................................... 58
3.3. Results .............................................................................................................................................. 58
3.3.1. Main Effects of Artificial Light at Night ..................................................................................... 58
3.3.2. Sensitivity Analyses ................................................................................................................... 59
3.3.3. Questionnaire 5 (2012-2013) .................................................................................................... 60
3.4. Discussion ......................................................................................................................................... 60
Chapter 4. Air Pollution and Risk of Non-Hodgkin Lymphoma in the California Teachers Study Cohort .. 84
4.1 Introduction ...................................................................................................................................... 84
4.2 Methods ............................................................................................................................................ 85
4.2.1 Study Population ........................................................................................................................ 85
4.2.2 Exposure Assessment ................................................................................................................. 86
4.2.3 Outcome Assessment ................................................................................................................ 88
4.2.4 Statistical Methods .................................................................................................................... 88
4.3 Results ............................................................................................................................................... 89
4.3.1 Harvard Air Pollution Model ...................................................................................................... 89
4.3.2 Kleeman Air Pollution Model ..................................................................................................... 90
4.3.3 Combined Light at Night and Air Pollution Model ..................................................................... 90
4.4 Discussion .......................................................................................................................................... 91
Chapter 5. Conclusion .............................................................................................................................. 106
Bibliography .............................................................................................................................................. 109

iv

List of Tables

Table 2.1 HLA associations with risk of DLBCL, FL, and CLL, reported based on HLA allelotyping data.
(Abbreviations: DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; CLL, chronic lymphocytic
leukemia; OR, odds ratio; RR, relative risk; NR, not reported) ................................................................... 32
Table 2.2 HLA (and loci on chromosome 6p21.3) associations with DLBCL, FL, CLL, and MZL, based on
imputation of SNP or genome-wide association data. (Abbreviations: DLBCL, diffuse large B-cell
lymphoma; FL, follicular lymphoma; CLL, chronic lymphocytic leukemia; OR, odds ratio; RR, relative risk;
NR, not reported) ........................................................................................................................................ 34
Table 2.3 Distribution of clinical stage for evaluated hematopoietic malignancies indicated for transplant
at the City of Hope at diagnosis .................................................................................................................. 38
Table 2.4 Distribution of transplant status for evaluated NHL malignancies (DLBCL, FL, and MCL) in
Caucasian patients indicated for transplant at the City of Hope ................................................................ 39
Table 2.5 Frequencies of HLA Alleles in City of Hope *Imputed Donors Compared to National Marrow
Donor Program ........................................................................................................................................... 40
Table 2.6 Distribution of hematopoietic malignancies indicated for bone marrow transplant at the City of
Hope 1995-2015 and potential donors (controls without cancer) for whom high resolution HLA
haplotyping was imputed ........................................................................................................................... 43
Table 2.7 HLA Haplotype association with B-cell NHL subtypes (DLBCL, FL, and MCL), in Caucasian
patients and donors (analyses adjusted for sex and age) ........................................................................... 44
Table 2.8 HLA Haplotype association with B-cell NHL subtypes (DLBCL and FL), in Hispanic patients and
donors (analyses adjusted for sex and age) ................................................................................................ 47
Table 2.9 HLA Haplotype association with DLBCL, in Asian patients and donors (analyses adjusted for sex
and age) ...................................................................................................................................................... 49
Table 2.10 HLA Haplotype association with CLL/SLL subtypes, in Caucasian patients and donors (analyses
adjusted for sex and age) ............................................................................................................................ 50
Table 2.11 Effect of homozygosity at the three HLA class I loci -A, -B and -C and three HLA class II loci -
DRB1, DQB1, and DPB1 on susceptibility to three B-cell NHL subtypes (DLBCL, FL, and MCL) in Caucasian
patients and donors (analyses adjusted for sex and age) ........................................................................... 51
Table 3.1 Correlation of US Defense Meteorological Satellite Operational Linescan System and Visible
Infrared Imaging Radiometer Suite Day/Night Band .................................................................................. 64
Table 3.2 Number of unique geocoded addresses in the California Teachers Study ................................. 65
Table 3.3 Demographic characteristics of the California Teachers Study Cohort in light at night analysis 66
Table 3.4 Associations between light at night and select demographic factors with non-Hodgkin
lymphoma ................................................................................................................................................... 68
Table 3.5 Additional sensitivity analyses for association between light at night and non-Hodgkin
lymphoma ................................................................................................................................................... 71
Table 3.6 Self-reported sleep quality and select baseline demographics among CTS participants reporting
same sleeping habits over the past year .................................................................................................... 73
Table 3.7 Association between self-reported sleep quality and artificial light at night among CTS
participants reporting same sleeping habits over the past year (n=40,613) .............................................. 75
v

Table 3.8 Association with light at night and non-Hodgkin lymphoma with follow-up beginning at time of
response to questionnaire 5 (2012-2013) .................................................................................................. 76
Table 4.1 Demographic characteristics of the California Teachers Study Cohort in air pollution analysis 94
Table 4.2 Association between particulate matter and non-Hodgkin lymphoma...................................... 96
Table 4.3 Comparison of Kleeman and Harvard air pollution models ........................................................ 98
Table 4.4 Association of Kleeman air pollution model and non-Hodgkin lymphoma ................................ 99
Table 4.5 Spearman correlation of particulate matter, World Atlas, and ultraviolet radiation ............... 100
Table 4.6 Multivariate model of light at night and air pollution and non-Hodgkin lymphoma risk in the
California Teachers Study ......................................................................................................................... 101

vi

List of Figures
Figure 3.1 Exclusion/inclusion criteria for California Teachers Study participants in analysis of artificial
light at night and cancer ............................................................................................................................. 77
Figure 3.2 Distribution of the New World Atlas of Artificial Sky Brightness in the California Teachers
Study cohort with quintile cutoffs .............................................................................................................. 78
Figure 3.3 Number of California Teachers Study Participants by County ................................................... 79
Figure 3.4 Median age at study entry of California Teachers Study participants by county ...................... 80
Figure 3.5 Number of non-Hodgkin lymphoma cases in the California Teachers Study cohort by county 81
Figure 3.6 Distribution of New World Atlas of Artificial Night Sky Brightness in the California Teachers
Study cohort ................................................................................................................................................ 82
Figure 3.7 Self-reported sleep questions from the California Teachers Study wave V questionnaire ....... 83
Figure 4.1 Exclusion/inclusion criteria for California Teachers Study participants in analysis of air
pollution and cancer ................................................................................................................................. 102
Figure 4.2 Change in Mean Concentration and Standard Deviation of PM2.5 (μg/m3) in the California
Teachers Study Cohort from 2000-2016 ................................................................................................... 103
Figure 4.3 Distribution of PM2.5 in the California Teachers Study cohort ............................................... 104
Figure 4.4 Distribution of PM0.1 in the California Teachers Study cohort ............................................... 105

vii

List of Abbreviations
AOD Aerosol optical depth
AP Air pollution
ALAN Artificial light at night
CARB California Air Resources Board
CLL Chronic lymphocytic leukemia
COH City of Hope
CTM Chemical transport model
CTS California Teachers Study
DLBCL Diffuse large B-cell lymphoma
DMSP United States Defense Meteorological Satellite Program
FL Follicular lymphoma
GWAS Genome-wide association study
HLA Human leukocyte antigen
IARC International Agency for Research on Cancer
IL Interleukin
Interlymph International Lymphoma Consortium
MHC Major histocompatibility complex
MZL Marginal zone lymphoma
NASA National Aeronautic and Space Administration
NDVI Normalized difference vegetation index
NHL Non-Hodgkin lymphoma
NMDP National Marrow Donor Program
NOAA National Oceanic and Atmospheric Administration
MAIAC Multi-Angle Implementation of Atmospheric Correction
MODIS Moderate Resolution Imaging Spectroradiometer
PAH Polycyclic aromatic hydrocarbons
PM Particulate matter
VIIRS Visible Infrared Imaging Radiometer Suite
WHO World Health Organization
WRF Weather Research and Forecasting

1

Chapter 1. Background
Non-Hodgkin lymphoma (NHL) is the most common form of cancer involving the lymph nodes,
comprising 80-90% of all lymphomas. It is one of the top 10 most common cancers in the United States
with an estimated 70,000 cases diagnosed in 2017(ACS, 2017). Although aged-adjusted incidence has
been largely stable over the past decade, the number of cases annually is predicted to rise to over
110,000 cases by 2030 as the population ages (Rahib et al., 2014). Survival has improved over the last
20 years, largely due to improved treatment, in particular the introduction of rituximab, a monoclonal
antibody that targets CD20+ B-cells. NHL is further subdivided into several subtypes based on cell of
origin and have been classified through the efforts of the International Lymphoma Consortium
(Interlymph) and World Health Organization (WHO) (Morton et al., 2007; Swerdlow et al., 2016; J. J.
Turner et al., 2010). The most common subtypes are diffuse large B-cell lymphoma (DLBCL, ~30%),
follicular lymphoma (FL, ~30%), and chronic lymphocytic leukemia/small cell lymphoma (CLL/SLL, ~12%).
This heterogeneous group of diseases has varying etiologies and outcomes with 5-year survival ranging
from 80% for indolent lymphomas such as FL, to less than 50% for more aggressive forms (ACS, 2017).
In addition to having heterogeneous clinical outcomes, recent population studies have supported the
presence of different etiologies and risk factors for these subtypes. For example, autoimmune
conditions increase risk for DLBCL and MZL but not follicular lymphoma; smoking has been reported to
increase risk for follicular lymphoma but not DLBCL; and hepatitis C and Helicobacter pylori are risk
factors for marginal zone lymphomas only (Morton et al., 2014). In contrast, some exposures such as
alcohol consumption appear to be associated (for this exposure, with decreased risk) uniformly across
all evaluated subtypes, suggesting some evidence of pleiotropy across the NHL subtypes. A role for
genetics has long-been supported by the association between family history of hematologic
malignancies and NHL risk. Genetic association studies thus far display similar patterns as non-genetic
risk factor association studies, demonstrating unique associations between genetic variations and NHL
2

subtypes, with some but little overlap. Of particular interest is that HLA has been implicated in every
NHL subtype evaluated to date. Like the environmental risk factor evaluations, however, individual
SNPs and HLA alleles have been identified as associated with the different NHL subtypes.
Few risk factors have been identified in lymphoma etiology, but as a cancer that originates in
the immune cells, factors that severely alter the immune response have long been suspected.
Laboratory, animal, and epidemiologic evidence have consistently supported a strong inflammatory and
immune component for lymphoma etiology (Smedby & Ponzoni, 2017). Established risk factors include
infection with the human immunodeficiency virus, inherited immunodeficiency syndromes (e.g., ataxia-
telangiectasia) and autoimmune conditions (e.g., Sjogren disease), organ transplant, and family history
of hematopoietic malignancies. Other immune-modifying risk factors include specific infections (e.g., H.
pylori for mucosa-associated lymphoid tissue (MALT) lymphoma of the stomach) and specific exposures
(e.g., breast implants and anaplastic large cell T-cell lymphoma). At present, confirmed lymphoma risk
factors explain only a minority of lymphomas. Although more common exposures such as obesity, some
pesticide exposures, and smoking have been reported, the strength of their association, if confirmed,
are modest at best. Recent epidemiologic efforts have comprised large-scale international pooled
analyses that have uncovered both environmental and genetic risk factors for lymphoma by subtype.
Immune dysregulation and chronic inflammation are recognized as a central factors in the
etiology of the leading causes of morbidity and mortality, including cancers, cardiovascular diseases, and
diabetes. Chronic inflammation, which reflects sustained activation of the innate and adaptive immune
systems is recognized as one of the hallmarks of cancer (Hanahan & Weinberg, 2011) and is recognized
as a major risk factor for a growing number of cancers (de Visser & Coussens, 2006). A growing number
of studies have linked immune markers from prediagnostic blood and subsequent cancer risk, adding
3

epidemiologic evidence to support the hypothesis that chronic inflammation is central to carcinogenesis
(Hosnijeh et al., 2016; Elena Vendrame et al., 2014; E. Vendrame & Martinez-Maza, 2011).
The pleiotropic nature of immune markers/cytokines makes the evaluation of each of these
responses pertinent to understanding the underlying biology of inflammation and cancer. The four T-
cell mediated immune responses include two main established patterns of immune response (Figure 1).
Th1 or “type 1” responses, are characterized by the production of the cytokine interferon-γ (IFNγ) upon
stimulation of CD4 T cells by IL-12, and are associated with enhanced anti-viral responses including the
stimulation of cytotoxic T cells (CTL), increased expression of human leukocyte antigen (HLA) class I cell
surface receptors, and natural killer (NK) cell activation. In contrast, Th2 or “type 2” responses are
characterized by production of IL-4, IL-5, and IL-13, and associated with enhanced antibody production,
especially those associated with allergic responses such as IgE. In the last decade, two more significant T
cell mediated immune responses have been identified, based on their patterns of cytokine secretion and
on the biological activities that are associated with them: (i) regulatory T cells (Treg) and (ii) Th17
response. Because each of the four T-cell mediated immune responses – Th1, Th2, Th17, and Treg – and
chronic inflammatory responses can contribute differentially to the carcinogenic process, identifying the
relative levels of circulating immune/cytokine markers could shed significant light on carcinogenesis.
4

Chapter 2. The Human Leukocyte Antigen and Non-Hodgkin Lymphoma Risk Among Transplant
Indicated Patients Seen at The City of Hope National Medical Center
2.1. Introduction
The human leukocyte antigens (HLA) is a class of genes encoded in the major histocompatibiltiy
complex (MHC) of chromosome 6 that is critical for human immune response. The MHC plays a major
role in immune response by presenting proteins on the cell surface to T cells. These genes encompass
an approximately 4 million base pair region at the 6p21.3 chromosome location. The HLA region is
highly polymorphic and each gene may have hundreds of different possible alleles leading to billions of
combinations among them. The HLA genes are broken down into classical and nonclassical groupings.
Among the classical HLA genes, it is the HLA Class I (HLA-A, HLA-B, and HLA-C) and HLA Class II (HLA-DP,
HLA-DQ, and HLA-DR) genes that are associated with antigen presentation. Classical HLA class I
receptors are expressed on every nucleated cell in the body. They present processed antigen to
cytotoxic T-lymphocytes (CD8
+
T-cells). HLA class II receptors are expressed only on professional
(immune) antigen-presenting cells including dendritic cells, monocytes/macrophages and B-cells; they
present antigen to helper T-lymphocytes (CD4+ T-cells). The CD8+ or CD4+ T-cell receptors form a link
with the HLA-bound antigen that initiates a cascade of events resulting in a cellular (cell-mediated) or
humoral (antibody) immune response (S. S. Wang & Hartge, 2010).
At present, HLA designation can be at the level of the gene, the HLA allele, the protein
expressed, or the antigen recognized by antibody. The identification of the HLA alleles began in the late
1950s and early 1960s, using serology to measure antibodies formed against the major HLA molecules.
Serological methods identified two major groups; group A which contained HL-A1, A2, A3, A9, A10, and
A11, and group B which contained HL-A5, A7, A8, A12, A13, A14, and A15. The World Health
Organization first convened the Nomenclature for Factors of the HLA System in 1968 to facilitate the
5

standardization of HLA nomenclature (Robinson et al., 2015). This workshop resulted in establishing
nomenclature whereby group A became HLA-A and group B became HLA-B, with the first two digits
corresponding to the allele group (i.e. HL-A7 became HLA-B*07). The workshop also created a
designation for HLA-Cw4, a broader epitope definition to denote antibodies that identified a different
antigen, but required additional typing to determine the specific allele. The designation “w” (for
workshop) was added to distinguish the major alleles from complement factors with the same initials.
With the introduction of sequencing technologies in the 1990s, HLA alleles could be delineated within
these broader groupings and Class II alleles (DR, DQ, and DP) were able to be resolved. Current
allelotyping methodologies now permit researchers to identify specific HLA receptor proteins that
correspond to specific HLA alleles (polymorphisms) and are denoted with nomenclature in a second set
of numbers (i.e. HLA-A*01:01) for the Class I (A, B, C) and Class II (DB, DR, DQ) alleles(Robinson et al.,
2015). Genetic sequencing techniques now permit researchers to correlate DNA polymorphisms to HLA
receptor (de Bakker et al., 2006).
The WHO Nomenclature Committee for Factors of the HLA System now presides over
nomenclature with a comprehensive overview provided at http://hla.alleles.org. As of June 2018,
13,680 HLA Class I alleles and 5,091 HLA Class II alleles were officially recognized by WHO(Robinson et
al., 2015). Each HLA allele receives an ID starting with the gene, followed by numerical codes that give
information on the allele group based on sequence homology to other alleles, next is the number of the
specific allele, followed by a number designation for the functional (changing an amino acid) nucleotide
substitution, next, a number for any nonsynonymous (nonfunctional) polymorphism, and finally a
number code for noncoding polymorphisms in an intron or untranslated region. For example, HLA-
DRB2*13:01:02 designates the gene encoding the DRBeta2 chain, *13 is the group of similar alleles, :01
designates a specific allele, and :02 defines a polymorphism that changes the amino acid residue in the
molecule that is different than that in HLA-DRB2*13:01:02:01.
6

Among the most polymorphic genes among humans, HLA alleles and haplotypes are increasingly
implicated in the etiologies and outcomes of immune conditions and hematopoietic malignancies,
including Hodgkin lymphoma (Johnson et al., 2015; McAulay & Jarrett, 2015), non-Hodgkin lymphomas
(NHL) (Abdou et al., 2010; Akers et al., 2011; Gragert et al., 2014; Y. Lu et al., 2011; Riemersma et al.,
2006; Rothman et al., 2006; C. F. Skibola et al., 2010; S. S. Wang et al., 2010; S. S. Wang et al., 2011), and
leukemia (El Ansary, Mohammed, Hassan, Baraka, & Ahmed, 2015; Naugler & Liwski, 2010; Orouji,
Tavakkol Afshari, Badiee, Shirdel, & Alipour, 2012; Speedy et al., 2014; Taylor et al., 2009). Recent
genome-wide association studies (GWAS) of NHL have now confirmed the role for HLA in four major NHL
subtypes: diffuse large B-cell lymphoma (DLBCL), follicular lymphoma (FL) etiology (CerhanBerndt, et al.,
2014; Conde et al., 2010; Smedby et al., 2011), chronic lymphocytic leukemia/small lymphocytic
lymphoma (CLL/SLL) (Berndt et al., 2013; S. L. Slager, Camp, et al., 2012; S. L. Slager et al., 2010), and
marginal zone lymphomas (MZL) (Joseph Vijai et al., 2015).
2.1.1. Diffuse Large B-cell Lymphoma
Initial clinic-based studies with relatively limited numbers of patients investigating the
association between HLA and NHL first implicated the tumor necrosis factor promoter (TNF)
polymorphism -308A, and specifically with DLBCL risk. Larger population studies that followed also
seemed to support the role of TNF-308A and DLBCL risk (Abdou et al., 2010; Nielsen et al., 2015; S. S.
Wang et al., 2007; S. S. Wang et al., 2009). A large pooled study of 1081 cases and 3564 controls within
the InterLymph consortium also reported the variant allele for TNF-308G>A (rs1800629) and
a TNF/LTA haplotype located in the MHC Class III region to be positively associated with DLBCL risk
(Rothman et al., 2006). Subsequent studies have extended that association to the ancestral haplotype,
AH 8.1, for which TNF is in linkage disequilibrium (S. S. Wang et al., 2011). The AH 8.1 comprises of HLA-
A*01:01~C*07:01~B*08:01~DRB1*03:01~DQA1*05:01~DQB1*02:01, and at 4.7 million nucleotides in
length, is the second longest haplotype known in humans. It is the most frequent extended haplotype
7

among Caucasian populations (15%) (Pociot et al., 1993; Wilson et al., 1993) and has been associated
with susceptibility to autoimmune and infectious diseases (e.g., celiac disease, Sjogren’s syndrome,
myasthenia gravis, type 1 diabetes) (Cameron, Mallal, French, & Dawkins, 1990; G. Candore et al., 1998;
Christiansen, Zhang, Griffiths, Mallal, & Dawkins, 1991) and with altered immune function (Giuseppina
Candore, Lio, Colonna Romano, & Caruso, 2002). Subsequent studies have attempted to clarify the role
of TNF-308A from AH 8.1, and from the individual HLA alleles that comprise AH 8.1. A recent study
implicating TNF-308A along with the AH 8.1 haplotype showed that the majority of individuals with the
AH 8.1 also had a variant TNF-308A allele (e.g., 99% in a U.S. Caucasian population) (S. S. Wang et al.,
2007), making it difficult to determine whether TNF-308A acts singularly without the effect of AH 8.1.
However, in that same population, the association between HLA-B*08 and DLBCL was more pronounced
than TNF or the AH 8.1 haplotype and the association between HLA-B*08 and DLBCL was evident
regardless of TNF G-308A status.
The evidence linking HLA B*08:01 to increased DLBCL risk is now consistent, replicated in
multiple studies, and confirmed in GWAS efforts (Tables 2.1 and 2.2) (Abdou et al., 2010; Nowak et al.,
2007; S. S. Wang et al., 2011). This association has been seen in pooled GWAS studies in both Caucasian
and East Asian populations which found a genome wide significant association with rs2523607, which is
highly linked to B*08:01 (r
2
=0.91) (Bassig et al., 2015; CerhanBerndt, et al., 2014).
Another component of AH 8.1, HLA-DRB1*03, has also been implicated in DLBCL risk. HLA
DRB1*03:01 (OR=1.55, 95% CI=1.02-2.36) has been associated with DLBCL in both Caucasian (Abdou et
al., 2010) and Korean populations (OR=3.4, p-value=0.003)(Choi et al., 2008). A SNP initially implicated
in follicular lymphoma risk, rs10484561 (which is in complete linkage disequilibrium (LD) with the HLA-
DRB1*01:01~DQA1*01:01~DQB1*05:01 extended haplotype) was also implicated in DLBCL risk in
another large-scale study of 4,449 NHL patients (OR=1.36, 95% CI=1.21-1.52) (Smedby et al., 2011).
8

Other associations with DLBCL that have been reported in Caucasian populations but not yet
validated in large population studies or GWAS include increased risk associations with DRB1*01
(OR=1.69, 95% CI=1.25-2.28) and decreased risk associations with DRB1*04:01 (OR=0.45) (S. S. Wang et
al., 2010) and DRB1*13 (OR=0.20) (Choi et al., 2008). Organ site-specific associations have also been
reported for DRB1*12 and increased risk of testicular DLBCL (OR=4.17, 95% CI=1.95-8.93) and with
DRB1*07 and nodular sclerosing DLBCL (OR=0.13, 95% CI=0.03-0.68) (Riemersma et al., 2006).
Increased DLBCL risk associated with B*51 has also been reported in a small study (43 cases, 200
controls) in both and Caucasian (OR=2.01) (S. S. Wang et al., 2010) and Korean populations (OR=3.0)
(Choi et al., 2008; González-Galarza et al., 2015).
Because allele frequencies between race groups are different – e.g., prevalence of AH 8.1 is
extremely low in other race/ethnic groups (<1%), we further consider divergent results that have been
reported in non-Caucasian populations. Coupled with the fact that DLBCL prevalence is also different by
race/ethic groups, it is plausible that there would be distinct HLA associations by different race/ethnic
groups. Studies among Korean populations have reported reduced DLBCL risk with HLA-A*33 (OR=0.37,
p-value=0.02) and HLA-B*44 alleles (OR=0.14, p-value=0.002)(Choi et al., 2008). Among a Chinese
population, the rs2647012 SNP in the HLA-DQB1 region was associated with increased DLBCL risk
(OR=1.28, 95% CI=1.02-1.61) (Qiao et al., 2013). Because of the small sample sizes of these studies,
however, we cannot rule out the possibility that these associations are false positive. Among larger
GWAS studies of Asians, Bassig and colleagues evaluated GWAS results identified in Caucasian
populations among 1124 DLBCL cases and 3596 controls reported genome-wide significant association
between HLA-B SNP rs25236078(B*08:01) and DLBCL among Eastern Asians (Bassig et al., 2015). Ten
and colleagues’ GWAS of 253 B-cell NHL and validation among 1,175 Chinese cases and 5,492 controls,
however, did not replicate HLA associations reported in Caucasian populations at a significant level for
DLBCL though similar directions in associations were noted (Ten et al., 2017).
9

Additional non-HLA SNPs found in the HLA region have also been associated with DLBCL. Most
notably, the rs6457327 SNP that codes for the Chromosome 6 Open Reading Frame 15 protein (C6orf15)
has been associated with reduced risk of DLBCL (Lim et al., 2014; C. F. Skibola et al., 2009). Nielsen et al.
also reported a decreased risk another TNF SNP (rs1799964) (Nielsen et al., 2015). Although Habermann
and colleagues reported a strong association between the transporter 2 ATP binding gene (rs241447)
and DLBCL risk (in a population study of 1193 NHLs) (DLBCL, OR=1.38, 95% CI=1.08-1.77) (Habermann et
al., 2008), results of TAP2 remain mixed with studies showing both increased risk (Cerhan et al., 2012)
and decreased risk (Nielsen et al., 2015).
An initial evaluation of HLA zygosity based on allelotyping data reported increased risk of DLBCL
with HLA class I zygosity among a U.S. Caucasian population of 163 DLBCL patients (S. S. Wang et al.,
2010). A GWAS-based evaluation of HLA zygosity which leveraged imputed HLA data with 3,617 DLBCL
patients confirmed the increasing FL risk with increasing number of HLA Class I alleles, but also reported
increasing DLBCL risk associated with HLA Class II alleles as well as zygosity across HLA-B, HLA-C, HLA-
DRB1, and HLA-DQB1 (Sophia S. Wang et al., 2018). However, among transplant recipients, Hussain and
colleagues reported no association among HLA Class I or II zygosity (Hussain et al., 2016).
Prognostic studies. As with etiologic studies, TNF polymorphisms and specifically TNF G-308A
has been extensively implicated in adverse NHL outcomes and specifically DLBCL prognosis (Juszczynski
et al., 2002; Seidemann et al., 2005; Krzysztof Warzocha et al., 1998; K. Warzocha et al., 2000; K.
Warzocha, Salles, Bienvenu, Barbier, et al., 1997; K. Warzocha, Salles, Bienvenu, Bastion, et al., 1997;
Younes & Aggarwall, 2003). Inferior progression free survival (Juszczynski et al., 2002; Nowak et al.,
2008) and inferior overall survival (Habermann et al., 2008; Juszczynski et al., 2002) have long-been
implicated for carriers of the TNF alpha promoter polymorphism G308-A and/or AH 8.1. These
associations have over time been more extensively evaluated in the context of HLA and other SNPs
10

within the 6p21.3 region. A U.S. cancer registry-based study of NHL by Lu and colleagues also reported
inferior overall survival in patients with HLA-C*07:01 (Y. Lu et al., 2011); whereas Bielska and colleagues
reported inferior FPS and OS with the B*13 and B*44 alleles, and inferior OS with HLA-G (Bielska et al.,
2015). In an analysis of HLA class I alleles in 154 NHL patients, HLA-B*08 and HLA-A*01 alleles were also
identified with poor NHL prognosis (Nowak et al., 2007). Additional associations include decreased 5-
year progression free survival and 5-year overall survival with HLA-B*44 and HLA-B*18 in a single center
study in Spain of 250 DLBCL patients (Alcoceba et al., 2013). Nielsen and colleagues’ study of 216 FL
patients further reported inferior OS with the TAP2 (rs241447) SNP. Nielsen and colleagues also
reported the TAP2 coding SNP rs241447 at 6p21.3 and inferior OS (Nielsen et al., 2015). Despite these
sporadic associations, a recent meta-analysis of four studies among Caucasian patients that evaluated
event-free survival, no SNPs in the HLA region or more broadly within chromosome 6p21.3 were
implicated (Ghesquieres et al., 2015).
2.1.2. Follicular Lymphoma
Three studies of HLA allelotyping among Caucasian populations have been conducted, ranging
from 163 to 265 FL patients (Table 2.1) (Akers et al., 2011; C. F. Skibola, Akers, et al., 2012; S. S. Wang et
al., 2010). Despite their relatively limited sample sizes, they have collectively implicated alleles in the
HLA Class II region. HLA-DRB1*01:01 and HLA-DQB1*05:01 have both been associated with
approximately a two-fold increased FL risk (Akers et al., 2011; Conde et al., 2010; C. F. Skibola, Akers, et
al., 2012; Christine F. Skibola et al., 2014; S. S. Wang et al., 2010) whereas HLA-DRB1*13 (OR=0.48, 95%
CI=0.28-0.82)(S. S. Wang et al., 2010), HLA-DQB1*06 (OR=0.51, 95% CI=0.39-0.69) (Akers et al., 2011),
and HLA-DRB1*15 (OR=0.53, 95% CI=0.34-0.81) and HLA-DPB1*03:01 (OR=0.39, 95% CI=0.21-0.68) (C. F.
Skibola, Akers, et al., 2012) have been associated with decreased risk of FL. Although the associations
between HLA-DRB1*01:01 and HLA-DQB1*05:01 with FL risk have been reported in multiple studies, the
remaining alleles have not been replicated among HLA allelotyping studies conducted to date.
11

A number of GWAS efforts among Caucasian populations have since been conducted, including
various single-center studies (Table 2.2) (Cerhan et al., 2012; C. F. Skibola et al., 2009; Wrench et al.,
2011), meta-analyses of these studies (C. F. Skibola, Conde, et al., 2012; Smedby et al., 2011), and an
international pooled GWAS effort comprised on population and clinic-based studies in the US, Canada,
and Australia comprising 1500 FL patients and ~7000 controls (Conde et al., 2010; Nieters et al., 2012;
Christine F. Skibola et al., 2014; Smedby et al., 2011). Notably, the most consistent association arising
from these studies is the two-fold association between the rs10484561 SNP and FL risk (Conde
combined p-value=1.12x10
-29
(Conde et al., 2010); Cerhan p-value=1.11x10
-88
(Cerhan et al., 2012)) which
was also confirmed by Wrench and colleagues in the UK in samples taken from the Barts and London
NHS Trust (n=218) (p-value=3.5x10⁻⁹) (Wrench et al., 2011). Sequencing studies have since identified
rs10484561 to be in complete linkage disequilibrium (LD) with the HLA-
DRB1*01:01~DQA1*01:01~DQB1*05:01 extended haplotype (C. F. Skibola, Akers, et al., 2012), which
further supports the earlier allelotyping reports implicating HLA-DRB1*01:01 with FL risk. Notably,
GWAS efforts in both European and Asian populations have implicated rs10484561 in increasing FL risk
(Cerhan et al., 2012; Conde et al., 2010; Ten et al., 2017; Wrench et al., 2011). This now consistent
association between rs10484561 and FL provides evidence that genetic variation in the MHC Class II
region is strongly associated with FL susceptibility and appear to implicate the extended haplotype that
includes HLA-DRB1*01:01~DQA1*01:01~DQB1*05:01. Although Conde and colleagues also reported
two-fold increased risk of rs7755224 (p-value=2.00x10
-19
) associated with FL, it is in complete LD with
rs10484561 (Conde et al., 2010).
GWAS have also identified the rs2647012 loci (on 6p21.3) to be significantly associated with
reduced FL risk. A three-stage genome-wide association study reported a combined OR=0.64, p-
value=2×10(-21)) (Smedby et al., 2011). Although located 962 bp away from rs10484561, they do not
appear to be linked (r
2
<0.1 in controls), and upon mutual adjustment, independent associations were
12

evident from both rs10484561 and rs2647012 (rs2647012:OR(adjusted) =0.70, 95% CI=0.67-0.78,
P(adjusted) = 4x10
-12
; rs10484561:OR(adjusted)=1.64, P(adjusted)=5x10
-15
). Through fine mapping,
Skibola and colleagues showed that a different haplotype was tagged by rs2647012, that carriers of the
DRB1*15:01~DQA1*01:02~DQB1*06:02 haplotype harbored the rs2647012 variant, though a modest
reduction in FL risk for rs2647012 was still observed after adjusting
for DRB1*15:01~DQA1*01:02~DQB1*06:02 (Christine F. Skibola et al., 2014). The association between
rs2647012 and FL was also reported in the Mayo clinic study of 238 Caucasians (Cerhan et al., 2012) with
a similar decrease in risk and in two other GWAS efforts conducted in 83 Malaysian and Chinese
cases(Ten et al., 2017) and a pooled analysis of 1428 cases from the US, Canada, and Australia (Smedby
et al., 2011). Although additional alleles and susceptibility loci have been reported, some, such as
rs9275517 were found to be in high linkage disequilibrium with rs2647012 and no longer associated
with FL after adjustment for rs2647012 (C. F. Skibola, Conde, et al., 2012). In follow-up studies, Foo and
colleagues suggested that the risk allele ‘C’ from rs10484561 tags haplotypes carrying the allele
encoding the high risk Phe amino acid, while the protective allele ‘T’ from rs2647012 tags haplotypes
carrying the allele encoding the low risk Ser residue (Foo et al., 2013).
A third loci implicated by GWAS studies is have also identified susceptibility loci in HLA Class II
region, rs6457327, to be associated with reduced FL risk (OR=0.59, 95% CI=0.50-00.70)(C. F. Skibola et
al., 2009), which sequencing studies subsequently implicated to be highly correlated with
DRB1*15:01~DQA1*01:02~DQB1*06:02 (Sillé et al., 2013), confirming results from earlier allelotyping
studies that have implicated HLA-DRB1*15 (C. F. Skibola, Akers, et al., 2012) and HLA-DQB1*06 (Akers et
al., 2011). Skibola and colleagues reported that C*07:02 and B*07:02 alleles were in LD with the
protective rs6457327 A allele (C. F. Skibola et al., 2009). However, individuals with rs6457327 had
approximately the same risk regardless of C*07:02 and B*07:02 status, suggesting the role for a yet
unidentified causal locus that is in LD with rs6457327. rs6457327 has also been consistently reported to
13

be associated with reduced FL risk in a pooled effort among cohort studies within the PAGE consortium
(OR=0.78, 95% CI=0.64-0.95) (Lim et al., 2014), and a clinic-based study with the Wellcome Trust Case-
Control Consortium (OR=0.75, 95% CI=0.61-0.93) (Wrench et al., 2011). rs6457327 overlaps C6orf15 and
is near psoriasis susceptibility region 1 (PSORS1); the risk allele (‘A’ allele) at rs6457327 is in high LD
with C*07:01 (D′ = 0.93) and B*07:02 (D′ = 1.0), though it remains to be confirmed whether these allele
account for the risk association (C. F. Skibola et al., 2009).
The decreased FL risk previously reported in allelotyping studies for DPB1*03:01 (C. F. Skibola,
Akers, et al., 2012) was further confirmed in a GWAS meta-analysis and follow-up allelotyping (C. F.
Skibola, Conde, et al., 2012). The GWAS initially identified the rs311722 to be associated with FL, but
noted that it is 6 kb downstream of HLA-DPB1 and is in linkage disequilibrium with HLA-DPB1*03:01.
Evaluation of rs311722 with DPB1*03:01 included in the same statistical model resulted in the
association between rs311722 and FL to be no longer significant. Skibola and colleagues’ meta-analysis
also implicated rs9275517 with follow-up allelotyping implicating the DQA1 regions and specifically,
DRB1*01:01~DQA1*01:01~DQB1*05:01 and FL risk (C. F. Skibola, Conde, et al., 2012). A pooled GWAS
effort further reported increased FL risks between DRA (OR=1.78, 95% CI=1.69-1.88), DRB1*01
(OR=1.85, 95% CI=1.70-2.03) and DRB1*07:01 (OR=1.52, 95% CI=1.39-1.66) with genome-wide
significance (Christine F. Skibola et al., 2014). Notably, the significant association with DRB1*01 and the
DRB1*01:01~DQA1*01:01~DQB1*05:01 haplotype is consistent with the earlier allelotyping-based
efforts by Wang and colleagues (S. S. Wang et al., 2010). Additional independent associations also
further implicate DQB1 in the association between rs17203612 and FL (OR=1.44, 95% CI=1.32-
1.57)(Christine F. Skibola et al., 2014) and rs2647046 and FL (OR=0.59, p-value=3.77x10
-10
), which are
not in LD with any of the previously identified SNPs (J. Vijai et al., 2013). Skibola also confirmed the
previously reported inverse associations between DQB1*06 and DRB1*13 using allelotyping studies;
although decreased FL risk was observed with DRB1*13 and DQB1*06, extensive linkage disequilibrium
14

with HLA-DRB1*15 and in general among DRB1, DQA1, and DQB1 loci make it challenging to pinpoint
the susceptibility loci (Christine F. Skibola et al., 2014).
Other implicated HLA-related SNPs which have not yet been replicated include rs241447 (TAP2,
OR=1.81, 95% CI=1.46-2.24) (Cerhan et al., 2012) and rs3132453 (PRRC2A, OR=0.59, 95% CI=0.48-0.73)
(Cerhan et al., 2012). TAP2 was reported by Cerhan and colleagues among 1193 NHLs and found to be
strongly associated with FL and was independent of other known loci (rs10484561 and rs2647012) from
this region (Cerhan et al., 2012).
To date, only two non-Caucasian GWAS of FL has been conducted. Ten et al conducted a small
hospital-based study of Malay and Chinese NHL patients and reported associations between two DQB1
snps (rs2647012 and rs10484561)(Ten et al., 2017). Specifically, the rs10484561 (OR=2.38, 95% CI=1.31-
4.34) was associated with a two-fold increased FL risk while the rs2647012 was associated with a two-
fold decreased risk (OR=0.56, 95% CI=0.33-0.94), both similar magnitudes of risk reported among
Caucasian population; however both were not significant after multiple testing correction. The study
provides the first evidence of replication of FL risk alleles identified in Caucasian populations in a non-
Caucasian population. Similarly, Tan and colleagues’ GWAS of 253 B-cell NHL and validation among
1,175 Chinese cases and 5,492 controls did not replicate HLA at a significant level though similar
directions in associations were noted for FL (Tan et al., 2013). Larger studies and studies among other
race/ethnic groups are required to substantiate these initial reports.
Only one report of zygosity has been reported to date for FL risk, reporting increased risk for FL
risk with zygosity in HLA Class II alleles, both overall across alleles (p-trend <0.0001) and individually
(DRB1: OR=1.54, 95% CI=1.31-1.82; DQB1: OR=1.42, 95% CI=1.23-1.65; DPB1: OR=1.24, 95% CI=1.10-
1.40)(Sophia S. Wang et al., 2018).
15

In a sample of the U.S. population of FL patients, HLA-A*01:01 was associated with inferior
overall survival while the HLA-Bw4 epitope and HLA-DRB1*13 were associated with better survival (Y. Lu
et al., 2011). The AH 8.1 haplotype was also associated with worse survival. In clinical studies of
Caucasians, the C6orf15 rs6457327 SNP was associated with increased risk of progression,
transformation of disease, and inferior survival (Berglund, Enblad, & Thunberg, 2011; C. F. Skibola et al.,
2009; Wrench et al., 2011).
Of the few studies of FL survival, the same C6orf15 SNP implicated in FL risk (rs6457327, risk
allele ‘A’ and associated genotypes ‘AA’ and ‘AC’) has been reported by studies conducted in Sweden
and the UK to be associated with risk for transformation from FL to DLBCL, predicting both time to (p-
value=0.02) and risk of (p-value=0.01) FL transformation, independent of clinical variables including the
follicular lymphoma international prognostic index (FLIPI) (Berglund et al., 2011; Wrench et al., 2011). A
prognostic study based on allelotyping data in the US reported inferior overall survival with HLA-A*01
and associated AH 8.1 haplotype as well as inferior OS was for HLA-Bw4 and HLA-DRB1*13 (Y. Lu et al.,
2011).
2.1.3. Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma
A number of modestly sized studies have investigated the role of HLA in CLL/SLL risk with a
growing consensus that HLA-DRB4*01 loci increase CLL risk (Table 2.1). A study of 101 German CLL
patients reported increased CLL risk with several class II alleles, including HLA-DRB4*01:03 (RR=2.74, p-
value=0.0025) (Machulla et al., 2001) , and a two-fold increased CLL risk with HLA-DRB1*04:01 (RR=2.13,
p-value=0.035), both of which are in linkage disequilibrium and potentially implicate CLL associations
with the larger HLA-DR4~DR53~DQ8 haplotype, which has also been related to susceptibility for several
auto-immune diseases (Machulla et al., 2001). Though not statistically significant after correction for
multiple comparisons, suggestive risk increase were also reported for HLA-DRB1*04:01, HLA-
16

DQB1*03:02 and HLA-DPB1*03:01. Another study of German patients also reported association with
HLA-DRB4 loci and CLL in all groups, gender, and age (Mueller & Machulla, 2002). Early studies
conducted by Dorak and colleagues also reported associations between the HLA-A2~B62~DR4 haplotype
with CLL (RR=4.1, p-value=0.002), and at the time also posited the link with HLA-DRB1*04:01 (RR=2.4, p-
value=0.009) (Dorak et al., 1996).
Notably, the HLA-DRB4*01:01 allele was also implicated in CLL risk in one of the largest studies
of HLA and cancer. In the study of nearly 3,500 CLL patients indicated for stem cell transplant conducted
through the National Marrow Donor Program and leveraging HLA typing data from 50,000 donor
controls (Table 2.1) (Gragert et al., 2014), the DRB4*01:01 allele and the
DRB4*01:01~DRB1*07:01~DQB1*03:03 haplotype was associated with a 1.5 fold increase in CLL risk
among Caucasians. It was also replicated in Hispanic (13.86-fold) and African American (28-fold)
populations; the unusually high magnitudes of risk are likely attributed by the smaller sample size
coupled with the lower frequencies of both DRB4*01:01 and the extended haplotype in non-Caucasian
populations (M. C. Di Bernardo et al., 2013) (Dorak et al., 1996).
Reported associations between HLA Class I alleles and CLL risk have been mixed. A 2006 study
of 98 patients in Spain found the HLA-Cw*06 allele was associated with a 2.69 increased risk of
developing CLL(Montes-Ares et al., 2006). A study of German patients also observed the association
with HLA-Cw*06, but only among early onset CLLs (Machulla et al., 2001; Mueller & Machulla, 2002).
Among CLL patients with a family history of CLL, the DRB1*11 alleles were associated with increased CLL
risk (p-value=0.009) (Theodorou et al., 2002). The study conducted among allogeneic transplant CLL
recipients also reported an association between HLA-A*01:01 and increased CLL risk and specifically
with the A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01 haplotype (OR=0.83, 95% CI=0.70-0.99),
17

and with A*02:01 (OR=1.2, 95% CI=1.08-1.34) and its haplotype
A*02:01~B*15:01~DRB1*04:01 (OR=1.41, p-value=0.0049) (Gragert et al., 2014).
Different associations have also been reported in different race/ethnic populations, including
increased risk associations between HLA-A*11:01 (OR=3.38, 95% CI=1.16-10.55) and HLA-B*35:01
(OR=3.51, 95% CI=1.48-8.6) and their combined haplotype (p-value=0.036), HLA-DRB1*07 (OR=2.19,
95% CI=1.02-4.75), DQB1*06 (OR=1.19, 95% CI=1.13-3.2), and DRB1*13~DQB1*03 haplotype (OR=4.29,
95% CI=1.29-15.75) among Iranian CLL patients (Mohammad Hojjat-Farsangi et al., 2008; M. Hojjat-
Farsangi, Razavi, Sharifian, & Shokri, 2014). In the same population, decreased CLL risk was reported for
HLA-A*01:01 (OR=0.34, 95% CI=0.13-0.87), HLA-A*26:01 (OR=0.32, 95% CI=0.10-0.92), HLA-B*65:01
(OR=0.17, 95% CI=0.03-0.91), HLA-B*53:01 (OR=0.04, 95% CI=0.01-0.32), and DQB1*03 (OR=0.58, 95%
CI=0.36-0.91). Among the allogeneic transplant population, increased CLL risk was also reported for
B*4:01 and C*07:04 among Hispanics, and DRB1*09:01 among African Americans. It is noted that some
of these associations, such as between HLA-C*04:01, may be contributed to their higher frequencies
among African American populations (Gragert et al., 2014).
Multiple GWAS efforts have been conducted with CLL, first among familial CLL subsets, and
more recently with all CLLs (Table 2.2). Curiously, initial reports of CLL GWAS studies identified multiple
risk loci, including 15q25.2, 2q37.1, 6p25.3, 11q24.1, 15q23, 19q13.32, 6p25.3, 8q24.1, 15q21.3, 16q4.1
though not initially in the HLA region (D. Crowther-Swanepoel et al., 2010; Crowther-Swanepoel et al.,
2011; Dalemari Crowther-Swanepoel et al., 2010; Maria Chiara Di Bernardo et al., 2008). Di Bernardo
and colleagues reported a CLL association with the HLA class I SNP rs6904029, which is also associated
with EBV + cHL, further supporting the association with A*02:01(M. C. Di Bernardo et al., 2013; Niens et
al., 2007). Recent GWAS efforts have implicated SNPs in the HLA-DRB1 (rs926070, OR=1.26)(Speedy et
al., 2014) and DQB1 (rs9273363, OR=1.24) regions(Berndt et al., 2013). A GWAS based on 102 patients
18

who reported a familial history of CLL with validation in 252 familial CLL cases reported a 1.87-fold
increased risk in the highly linked DQA1(rs602875) and DRB5(rs674313) regions (combined P-
value=6.92x10
-9
) (Susan L. Slager et al., 2011); because imputation of HLA alleles in this region was not
performed it is unclear which specific loci these SNPs are tagging. Additionally, Slager and colleagues
genotyped 1196 CLL cases and 2410 controls and found significant associations between two variants in
6p21.31 variants and CLL risk (rs210134: p-value=0.01; rs210142: p-value=6.8×10(-3)). rs210134 is in
intron 1 and is 107 bp from rs201142 (S. L. Slager, Camp, et al., 2012). In a meta-analysis of 3 genome-
wide association studies based on a discovery phase of 1121 cases and 3745 controls with replication
analysis in 861 cases and 2033 controls. CLL risk locus at 6p21.33 (rs210142; intronic to the BAK1 gene,
BCL2 antagonist killer 1; p-value=9.47×10(-16)) were reported (S. L. Slager, Skibola, et al., 2012). A
strong relationship between risk genotype and reduced BAK1 expression was shown in lymphoblastoid
cell lines.
A handful of studies have explored the possible link between HLA zygosity and CLL risk. One
large-scaled pooled effort of GWAS based on imputation methodologies for HLA reported modest but
statistically significant associations for CLL risk among 2878 patients and 8753 controls with
homozygosity at HLA-A (OR=1.19), HLA-DRB1 (OR=1.19), and HLA–DQB1 (OR=1.2) (Sophia S. Wang et al.,
2018). Notably, similarly modest associations were also reported independently by Gragert and
colleagues in the transplant eligible CLL cases from the National Donor and Marrow Program, where
statistically significant odds ratios of 1.19 were reported for homozygosity at HLA-A (Gragert et al.,
2014). Modest CLL risks were also reported for homozygosity at HLA-A (OR=1.19) and for HLA-B and
HLA-C (ORs=1.16). The increased CLL for HLA-DQB1 homozygosity has also been reported in smaller
populations, including a German CLL patient population (Machulla et al., 2001). Another population of
German CLL patients also reported increased CLL risk with HLA-DQB1 homozygosity, but their
association was specific for female patients(Mueller & Machulla, 2002). Mueller and colleagues’ report
19

also supported the increased CLL risk with DRB3/4/5 loci, but again the association was more
pronounced among female patients than male CLL patients. Another small study of 79 CLL patients and
329 controls in Germany reported DR53 homozygosity to be associated with CLL (RR=2.4, p-value=0.03),
with risk to be even more pronounced among early onset of CLL (RR=4.4, p-value=0.008) (Dorak et al.,
1996).
Among the number of studies that have investigated genetic susceptibility to CLL prognosis,
relatively few modest-sized studies of CLL prognosis have implicated HLA alleles. A study by Garcia-
Alvarez et al in a subset of 156 Spanish CLL-like monoclonal B-cell lymphocytosis patients, specifically in
128 cases with IGHV mutation, reported a 2.3-fold increased risk of progression to CLL among those with
the DQB1*03 allele (García-Álvarez et al., 2017). They also observed increased treatment free survival
among those with DQB1*02 (p-value=0.012). HLA-C2 was associated with decreased PFS (49%±9% vs
75%±7%, p-value=0.02) in a Polish cohort of 197 cases (Karabon et al., 2011). Among Iranian patients,
significant risk increase for disease progression was reported for DRB1*04 and DRB5 alleles, but they did
not identify differences in HLA-A or HLA-B alleles or haplotypes(Mohammad Hojjat-Farsangi et al.,
2008). On the other hand, Lech-Maranda and colleagues reported HLA-DRB1*01 (p-value=0.007) and
HLA-DRB1*02-null (p-value=0.002) to be associated with shorter overall survival (Lech-Maranda et al.,
2007). One study of the non-classical HLA loci suggested that the presence of HLA-E*01:03 allele and
detection of soluble HLA-E as an independent predictor of disease progression (Wagner et al., 2017).
Finally, Slager and colleagues evaluated the relationship between the putative 6p21.31 SNP (rs210142)
in a case-only analyses of CLL patients of Rai stage and platelet counts and found that CLL patients with
different alleles had different disease grades (major allele with stage 4 disease) the A allele with the
major allele had higher grade disease(S. L. Slager, Camp, et al., 2012). Furthermore, because CLL/SLL
patients with mutated immunoglobulin heavy chain variable region genes (IGHVs) have better clinical
outcomes, future prognostic studies will need to take IGHV hypermutation status into consideration.
20

2.1.4. Marginal Zone Lymphoma
Only one GWAS of MZL has been conducted to date. Among a European population of 1346
MZL cases, Vijai and colleagues reported significant genome-wide increased MZL risk with HLA SNP
rs2922994 (B*08:01, OR=1.64, 95% CI=1.39-1.92) and rs9461741 (OR=2.24, 95% CI=1.64–3.07), located
in HLA Class II in the intron between exons 3 and 4 of the BTNL2 gene and posited to be linked to HLA-
DRB1*01:02 (Joseph Vijai et al., 2015).
Only one study to date has reported associations between HLA zygosity and MZL, leveraging
GWAS data from 741 MZL patients that were part of the international GWAS effort and imputed for
HLA. Wang and colleagues reported that increasing number of HLA class I and class II loci was
associated with increased MZL risk, with specific MZL risk increases noted homozygosity at HLA-B
(OR=1.34, 95% CI=1.01-1.78), HLA-C (OR=1.33, 95% CI=1.04-1.70) and HLA-DRB1 (OR=1.45, 95% CI=1.12-
1.89) loci (Sophia S. Wang et al., 2018).
The studies in which associations between HLA loci and hematopoietic malignancies have largely
been conducted are epidemiologic and clinical studies, which depend on voluntary participation among
the catchment population. As a consequence, there are well-documented survival biases among these
studies where participation is understandably higher among patients who have less severe disease
(Cerhan et al., 2011; Holly, Gautam, & Bracci; Z. H. Hu, Connett, Yuan, & Anderson, 2016). Most
epidemiologic and association studies of NHL thus possess little to no representation of patients for
whom bone marrow transplants are indicated. We thus hypothesize that identifying risk alleles in this
population may provide insight into a more severe disease etiology. Our efforts here build upon
previously published results, including efforts based on the National Marrow Donor Program (NMDP)
which implicated HLA alleles and haplotypes in the risk of chronic lymphocytic leukemia (CLL) in their
transplant-based population (Gragert et al., 2014), and GWAS studies which have evaluated HLA allele
21

associations based on imputed data among epidemiologic and clinical studies (Cerhan et al., 2011;
Smedby et al., 2011; Joseph Vijai et al., 2015). Although the NMDP effort similarly used HLA typing data
to identify loci for a transplant population of CLL, they were unable to evaluate other NHL subtypes due
to their lack of subtype information within the NMDP. Here, we link pathology records, HLA typing, and
patient data from a single institution’s (City of Hope) transplant database to evaluate the associations
between HLA alleles and haplotypes with NHL subtypes.
2.2. Methods
2.2.1. Study Population
Our patient population comprised individuals diagnosed with NHL at the City of Hope (COH)
between January 1995 and October 2015 and indicated for a transplant. To be included in the analysis,
patients were identified with the COH pathology database and medical records and then linked to the
COH HLA database. Prospective donors who were typed for HLA at COH during the same time period
served as the comparison/control group. A total of 1,977 patients with histologically confirmed non-
Hodgkin lymphoma (NHL) per Interlymph/WHO classification (Morton et al., 2007; J. J. Turner et al.,
2010) and 15,020 donors were identified between January 1995 and October 2015.
2.2.2. NHL definition
NHL subtypes were defined according to the WHO algorithm. Briefly, DLBCL was defined as ICD-
O-3 codes 9678, 9679, 9680, and 9684; FL was defined as 9690, 9691, 9695, and 9698; CLL/SLL was
defined as 9823 and 9670; mantle cell was defined as 9673 (J. J. Turner et al., 2010). As expected, NHL
patients indicated for transplant had more advanced disease than the general COH population (Table
2.3). Of the patients indicated for transplant and typed for HLA, 30% did not receive a transplant; and of
those who did, the majority received an autologous transplant (Table 2.4).
22

To evaluate the comparability of the COH HLA donor population to that of the general
population, we compared HLA allele frequencies (minimum 1% frequency) to previously published
frequencies (González-Galarza et al., 2015) and noted the similarity, including with the NMDP
population (Table 2.5).
2.2.3. HLA Typing and Imputation
HLA typing from 1995-2000 in the COH histopathology laboratory comprised a combination of
serological, low resolution, and high resolution HLA typing. All HLA typing after May 2000 was
conducted with low or high resolution molecular typing.
Briefly, for low to intermediate resolution HLA typing, a reverse sequence-specific
oligonucleotide (rSSO) probe hybridization method was utilized. Sequence-specific oligonucleotide
probes (SSO) bind to fluorescently coded microspheres to identify alleles encoded by the sample DNA.
The target DNA is amplified by polymerase chain reaction (PCR) and then hybridized to probes that are
specific to the HLA region. Signal detection is captured on a Luminex® 100 analyzer. Based on the
Luminex® xMAP® Technology, the system enables multiplexing of up to 100 analytes in a single
microplate well. Each well represents one sample and one HLA locus. Analysis is performed using the
HLA Fusion software from One Lambda. The program matches the positive and negative signals from
each probe to a reaction pattern that defines a specific HLA allele group and assigns HLA type.
For high resolution HLA typing, two different methods were used. The first method utilized was
the Sanger sequence-based typing method (SBT, SBT Resolver, Conexio Genomics) whereby the target
sequence is amplified, followed by treatment with ExoSAP-IT® to remove unincorporated primers and
deoxynucleotide triphosphates (dNTPs). The amplicon is then used as template for direct automated
fluorescent DNA sequencing using customized sequencing primers and the BigDye® Terminator
sequencing chemistry (Applied Biosystems™, Life Technologies™). The extension products are purified
23

according to the ethanol precipitation method and denatured using Hi-Di™ formamide (Applied
Biosystems™, Life Technologies™) before separation and detection on an automated fluorescent DNA
sequencer. Results are imported and analyzed using Assign™ sequence analysis software from Conexio
Genomics. The second method used to obtain high resolution HLA typing was next generation
sequencing (NGS) from Scisco Genetics. The method uses the Illumina MiSeq system, which is based on
a sequencing-by-synthesis approach utilizing fluorescently labeled reversible terminator nucleotides.
After assay specific amplification, samples are tagged with unique indexes and pooled together and
applied to the MiSeq instrument, where they are amplified as individual clusters and sequenced using
universal sequencing primers. The result is several million reads that can be analyzed using the GeMS
HLA software (Scisco Genetics) to report unambiguous HLA allele types for several individual samples
simultaneously.
To maximize four-digit HLA data and haplotype information for analysis, race, and HLA typing
data were used to impute high resolution alleles and haplotypes when possible. Briefly, given a
minimum of low resolution, 2 digit A-B-DR alleles and race, the possible list of high resolution A-B-C-DR-
DQ haplotypes is enumerated though an expectation-maximization algorithm (Gourraud, Lamiraux, El-
Kadhi, Raffoux, & Cambon-Thomsen, 2005). The probability of a given haplotype is calculated based on
registry level data (Gragert, Madbouly, Freeman, & Maiers, 2013; Madbouly et al., 2014; Paunić,
Gragert, Madbouly, Freeman, & Maiers, 2012). We generated 5 unambiguous imputed HLA datasets
based on these probabilities for analysis.
2.2.4. Analytic Population
Of the 1,977 NHL patients indicated for transplant and 15,020 controls; age, race, and A-B-DR
HLA genotyping was available for 1,366 patients and 10,271 donors, and who comprised the final
analytic population. Many cases that underwent autologous transplant had insufficient HLA typing. City
24

of Hope historically performed low resolution A and B typing and anti-HLA antibody testing for all
autologous transplant patients in order to have initial typing in case of allogeneic transplant and
transfusion support. Of NHL patients with imputed haplotypes, there were 354 diffuse large B-cell
lymphomas (DLBCL), 263 follicular lymphomas (FL), and 173 mantle cell lymphomas (MCL).
2.2.5. Statistical Analysis
We conducted haplotype- and allele-level HLA analyses for four NHL subtypes (DLBCL, FL,
CLL/SLL, and mantle cell lymphoma). We calculated the risk (odds ratios) with each NHL subtype in
logistic regression models, adjusting for age and sex (Klein, Jedlicka, & Pekosz, 2010). Geometric mean
odds ratios were calculated after adjustment with the false discovery rate (FDR) method at a threshold
of 0.05 (Benjamini & Hochberg, 1995). Given the high linkage disequilibrium (LD) present in the HLA
region and co-dominant fashion in which HLA alleles are expressed, results are presented by haplotype.
Due to the high number of haplotypes, we restricted our analysis to haplotypes with at least a frequency
of one percent among donors. We further present results for the individual HLA loci affiliated with the
HLA haplotypes to determine whether any allele was the dominant association. We also evaluated the
association between HLA homozygosity for HLA Class I (A, B, C) and Class II (DR, DB, DQ) alleles, whereby
ORs and 95% CIs were calculated as estimates of NHL risk with heterozygotes as the referent category,
adjusted for sex, age, study, and race. Estimates for each additional number of homozygous loci were
calculated as p-trend. For all results, results stratified by race are presented. All analyses were
conducted in SAS 9.4 (SAS Institute, Cary, NC).
2.3. Results
Of the 1,366 NHL patients included in this analysis, 938 were Caucasian, 237 Hispanic, 138 Asian,
and 54 African American (Table 2.6). The most common NHL subtype was DLBCL (26%), followed by FL
(19%), mantle cell (13%), and CLL/SLL (11%). Overall, there were more male patients (63.5%) than
25

female patients. Among the donors, 6,801 were identified as Caucasian, 2,076 as Hispanic, 1,029 as
Asian, and 365 as African American.
Due to sample size, we present results for HLA associations with DLBCL, FL, and mantle cell
lymphoma (Table 2.7). Because COH participants are in the National Marrow and Donor Program, data
from the majority of our CLL patients were included in a previous publication on HLA and CLL
associations based on data at the NMDP (Gragert et al., 2014). We therefore present our institution-
specific results for CLL in Table 2.10. For all NHL subtypes, results for Caucasians are shown in the main
tables and those for non-white race groups, when sample size permits, are presented in Tables 8 and 9.
2.3.1. Diffuse Large B-cell Lymphoma
Among Caucasians, the most common haplotype in donors was the AH 8.1 haplotype (HLA-
A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01) at 7.2% (Table 2.7), but was not statistically
significantly associated with DLBCL risk (OR 1.19, 95% CI 0.66-2.15). We identified three haplotypes,
HLA-A*03:01~C*04:01~B*35:01~DRB1*01:01~DQB1*05:01 (OR 2.70, 95% CI 1.21-6.01) HLA-
A*02:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01(OR 3.27, 95% CI 1.46-7.32) and HLA-
A*26:01~C*12:03~B*38:01~DRB1*04:02~DQB1*03:02 (OR 3.06, 95% CI 1.34-6.97), that appeared to
confer an increased risk for DLBCL. Evaluation of the individual alleles which comprise those haplotypes
did not yield any significant associations with DLBCL risk. However, we observed two individual HLA
alleles which were significantly associated with increased DLBCL risk after FDR correction: HLA-B*13:02
(OR 1.99, 95% CI 1.21-3.26) and HLA-C*06:02 (OR 1.72, 95% CI 1.24-2.40). Both loci are part of the HLA-
A*30:01~C*06:02~B*13:02~DRB1*07:01~DQB1*02:01 haplotype, but that haplotype was not
significantly associated with DLBCL (OR 1.36, 95% CI: 0.46-4.10).
The AH 8.1 haplotype was only seen in 3% of Hispanic donors and 0.5% of Asian donors and not
associated with DLBCL risk (Tables 2.8 and 2.9). The most common haplotype in Hispanic donors was
26

HLA-A*68:02~C*08:02~B*14:02~DRB1*01:02~DQB1*05:01 (4%), but, while elevated, was not
statistically associated with DLBCL (OR 2.29, 95% CI 0.65-8.06). The most common haplotype in Asian
donors was HLA-A*33:03~C*03:02~B*58:01~DRB1*03:01~DQB1*02:01 (9%) and also not associated
with DLBCL (OR 1.48, 95% CI 0.18-12.48). No other haplotypes or individual-level alleles were associated
with DLBCL risk after correction for FDR among our Hispanic or Asian populations.
2.3.2. Follicular Lymphoma
Among Caucasians, the HLA- A*02:01~C*06:02~B*13:02~DRB1*07:01~DQB1*02:01 haplotype
was associated with increased FL risk (OR 2.70, 95% CI: 1.09-6.66) (Table 2.7). We also observed a few
haplotypes that were borderline significant, including HLA-
A*29:02~C*16:01~B*44:03~DRB1*07:01~DQB1*02:01 (OR 2.01, 95% CI: 0.96-4.22) and HLA-
A*23:01~C*04:01~B*44:03~DRB1*07:01~DQB1*02:01 (OR 2.69, 95% CI: 0.95-7.58). Evaluation of the
individual HLA alleles which comprise these haplotypes identified significant risk associations with HLA-
DRB1*07:01 (OR 2.41, 95% CI: 1.75-3.31) and HLA- DQB1*02:01 (OR 1.63, 95% CI: 1.16-2.30), both of
which remained statistically significant after FDR correction (Table 2.7). Among Hispanics, we observed
an increased risk for FL with the haplotype HLA- A*03:01~C*04:01~B*35:01~DRB1*01:01~DQB1*05:01
(OR=8.71, 95% CI: 2.06-36.88), with significant associations present for the HLA- DRB1*01:01 (OR 4.83,
95% CI: 2.39-9.78) and HLA -DQB1*05:01 (OR 2.65, 95% CI: 1.35-5.79) alleles (Table 2.8). HLA
associations for FL among Asians could not be evaluated as we only had 18 Asian FL patients with full
HLA haplotype information.
2.3.3. Marginal Zone Lymphoma
The HLA-A*02:01~C*03:04~B*15:01~DRB1*04:01~DQB1*03:02 haplotype was borderline
significantly associated with increased MCL risk (OR 3.13, 95% CI 1.00-9.83) among Caucasians, but no
individual alleles were significantly associated with MCL (Table 2.7). HLA-A*01:01 was associated with
27

MCL (OR 1.60, 95% CI 1.05-2.43), remaining significant after FDR correction, but the affiliated haplotype
was not associated with MCL. Despite the relatively small number of Asians in our study population, we
observed a strong association between the HLA-A*24:02~C*07:02~B*07:02~DRB1*01:01~DQB1*05:01
haplotype and MCL (OR 11.7, 95% CI 2.25-60.9, Table 2.9); the HLA-B*07:02 (OR 8.41, 95% CI 2.20-32.2)
and HLA DRB1*01:01 (OR 11.12, 95% CI 2.50-49.4) loci both remained statistically significant after FDR
correction. MCL risk could not be calculated due to the small number of Hispanic MCL patients in our
study population.
2.3.4. HLA Class I and II homozygosity
Among our Caucasian patients, we did not observe statistically significant associations between
overall HLA class I homozygosity and NHL subtype risk, though the associations were trending towards
increased risk with increasing homozygosity (Table 2.11). We observed increased DLBCL and FL risk with
homozygosity at HLA Class II loci, including HLA-DRB1 (DLBCL OR 1.89, 95% CI 1.28-2.79; FL OR 2.06 95%
CI 1.37-3.09) and HLA-DQB1 (DLBCL OR 1.46, 95% CI 1.04-2.06; FL OR 1.50 95% CI 1.04-2.16). Among FL,
we further observed an association between increasing number of homozygous HLA class II loci and
increased FL risk among Caucasians (p-trend = 0.03).
2.4. Discussion
Our analysis of HLA variation and NHL risk among a transplant-indicated patient population for
NHL confirmed previously identified risk alleles and identified some potentially novel alleles that
warrant follow-up and replication. Specifically, statistically significant associations, even after FDR
correction, were observed between HLA-DRB1*07:01 and FL among our Caucasian patients, which was
previously reported in GWAS efforts (Christine F. Skibola et al., 2014). Associations between HLA-
DRB1*01:01 and HLA-DQB1*05:01 and FL was also evident in our Hispanic patients (C. F. Skibola, Akers,
et al., 2012; S. S. Wang et al., 2010). For DLBCL, significant association between the HLA-
28

A*02:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01 haplotype was observed, which notably shares
the same previously GWAS-implicated HLA-B*08:01~DRB1*03:01 region that is also in high LD with the
previously reported risk allele, TNF-308A, as well as the HLA-DR3-DQ2 serotype that is associated with
several other autoimmune conditions (Smigoc Schweiger et al., 2016). Individually, however,
associations between GWAS-confirmed loci, HLA-B*08:01 and HLA-DRB1*03:01, and the ancestral
haplotype 8.1 (HLA-A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01) were null in our population
(Abdou et al., 2010; CerhanBerndt, et al., 2014).
The other two HLA haplotypes we observed with increased DLBCL risk in our Caucasian
population (HLA-A*26:01~C*12:03~B*38:01~DRB1*04:02~DQB1*03:02 and HLA-
A*03:01~C*04:01~B*35:01~DRB1*01:01~DQB1*05:01) have not previously been reported and require
replication. We note that the HLA-A*26:01~B*38:01~DRB1*04:02 haplotype is more commonly seen in
Ashkenazi Jewish populations (2.3-6.7% frequency)(Klitz et al., 2010; Manor et al., 2016) compared to
other non-Hispanic European populations (<0.5% frequency) (González-Galarza et al., 2015) and it is
possible that this association may reflect a bias in our case population. We also observed increased
DLBCL risk with the HLA-B*13:02 and HLA-C*06:02 alleles, even after FDR correction, the latter which is
implicated in increased risk of CLL (Gragert et al., 2014) and psoriasis (Guðjónsson et al., 2002) though
the two alleles are in high LD (D’=0.98). HLA haplotype and allele frequencies varied widely among
Hispanics and Asians, compared to Caucasians; the associations we observed among Caucasians was not
replicated among either Hispanic or Asian subgroups, though our sample size was admittedly limited.
In our Caucasian population, we replicated previously reported GWAS association for FL (HLA-
DRB1*07:01) (Christine F. Skibola et al., 2014) and further implicate the HLA-
A*02:01~C*06:02~B*13:02~DRB1*07:01~DQB1*02:01 haplotype as a possible FL risk haplotype. In
addition, we observed increased risk with the HLA-DQB1*02:01 allele which was in high LD with HLA-
29

DRB1*07:01 (D’ = 0.76) The replication of GWAS results is not entirely unexpected, as FL is in general a
more indolent disease for which minimal survival bias would be expected among population-based
studies. Of the NHL subtypes evaluated, the difference between our clinical subset of transplant-
indicated FL patients and that of previously published population studies would be the least dissimilar
(Cerhan et al., 2011). These associations were not, however, replicated in the other race groups.
Among Hispanics, we observed a very different haplotype association between HLA-
A*03:01~C*04:01~B*35:01~DRB1*01:01~DQB1*05:01 and FL. Although we cannot exclude the
possibility that HLA associations differ by race/ethnicity, we also cannot exclude the possibility that
these associations are false positive results due to the relatively small sample size available for analysis
in non-Caucasian race groups.
In mantle cell lymphomas, we found no significant HLA haplotype association, although we did
observe an association with the HLA-A*01:01 allele, an association previously reported for both FL (Y. Lu
et al., 2011) and EBV+ HL (Johnson et al., 2015). HLA-A*01:01 had been implicated as a risk factor in
type 1 diabetes but recent studies have attributed this to the risk associated with the AH8.1 haplotype,
and specifically to its linkage with the HLA-DRB1*03:01 allele (Noble, 2015). Although we observed no
association between the AH8.1 haplotype and MCL, the HLA-DRB1*03:01 allele was significantly
associated with increased MCL risk but did not remain significant after FDR correction. We observed
different associations among Asians in our study population, specifically associations between the HLA-
A*24:02~C*07:02~B*07:02~DRB1*01:01~DQB1*05:01 haplotype and MCL. Similar to the observations
observed for DLBCL and FL, our lack of replication of observed results in Caucasians coupled with new
distinct significant results among other race groups, suggests either that our results among other race
groups may be due to chance or a larger implication that distinct HLA associations exist for different
race/ethnic groups. We also note that MCL patients may also undergo transplant as part of
consolidation therapy and as such may not necessarily be a higher risk population compared to our
30

DLBCL or FL patients. At time of transplant, 51% of the MCL patients were in first complete response,
compared to only 10% of DLBCL and 7% of FL patients.
Our evaluation of HLA homozygosity was not entirely consistent to previous reports, which were
largely derived from population-based studies (S. S. Wang et al., 2010; Sophia S. Wang et al., 2018). We
did not observe a statistically significant overall increase in disease risk with increasing homozygosity at
the HLA class I loci. However, we did observe associations with HLA class II homozygosity, specifically
between HLA-DRB1 and HLA-DQB1 homozygosity and both FL and DLBCL risk (Sophia S. Wang et al.,
2018). Our data thus appear to support the notion that class II heterozygosity provides greater ability
for the host to enhance tumor surveillance in its role in antigen presentation (S. S. Wang & Hartge, 2010)
(Blackwell, Jamieson, & Burgner, 2009) and in clearing related infections (Thio et al., 2003) (De La
Concha et al., 1999). At present, it remains unclear what the underlying biology is, should one exist, for
the observed associations between the two specific HLA Class II zygosity and FL and DLBCL.
Strengths of our study include a relatively large sample size and availability of subtype
information for NHL. Our HLA typing was largely ascertained through genotyping, in contrast to GWAS-
based analyses which are imputation-based. The available of HLA typing information was critical in
permitting us to impute HLA haplotype information for data analysis. In addition, our comparison group
of donors was drawn from among individuals typed as possible matches for our patients. Because
donors include some family members, it is plausible that the comparison donor population might be
slightly more likely to have shared alleles, which would likely attenuate our results. However, we did
not see large differences in allele frequencies between our donors and the NMDP population. Study
strengths also include the unique study population of transplant-indicated NHLs, which differ from those
reflected in published HLA-association literature. In particular, our patients are diagnosed at higher
stage, as would be expected of those undergoing transplantation. Study limitations include the
31

exclusion of patients and donors for whom we did not have age, sex, and race. In addition, while our
sample size is relatively large by subtype, the variety of HLA haplotypes and alleles limit our statistical
power. We also note that a number of patients that underwent autologous transplant had insufficient
HLA typing for inclusion in our analysis; this is largely due to historical data at City of Hope for which only
low resolution A and B typing and anti-HLA antibody testing was available. While City of Hope performs
the most transplants for NHL in the nation ("Center for International Blood and Marrow Transplant, a
contractor for the C.W. Bill Young Cell Transplantation Program operated through the U. S. Department
of Health and Human Services, Health Resources and Services Administration, Healthcare Systems
Bureau. U.S. Transplant Data by Center Report, NHL - Non-Hodgkin Lymphoma, Number of Transplants
Reported in 2016. U.S. Transplant Data by Center Report. Last Updated: May 17, 2017.,"), our results
only represent a subset of patients seen at a single institution and will require replication in larger
populations.
To our knowledge this and the previous publication on CLL (Gragert et al., 2014) are among the
first evaluations of HLA association in NHL subtypes specifically among transplant-indicated populations.
Similar to the report of CLL, some GWAS associations reported among the general population were
replicated in this unique transplant-indicated population, but our results further suggest new
associations among more severe disease which require replication in other similar study populations.
Complementary analyses of these and other risk factors (genetic and non-genetic) for more severe
disease warranting transplant would shed light on whether there are unique risk factors or risk factor
combinations which indeed result in more aggressive disease.
32

Table 2.1 HLA associations with risk of DLBCL, FL, and CLL, reported based on HLA allelotyping data. (Abbreviations: DLBCL, diffuse large B-cell
lymphoma; FL, follicular lymphoma; CLL, chronic lymphocytic leukemia; OR, odds ratio; RR, relative risk; NR, not reported)

Subtype
Study description
(Reference)
Year
published Country
Race /
ethnicity
# cases /
# controls Loci
OR / RR
(95% CI) p-value
DLBCL
U.S. population-based
case-control study 2010 USA
Non-
Hispanic /
European 163 cases HLA class I
3.66
(1.15-11.7) 0.01

(Wang et al)

555 controls

International pooled
GWAS effort,
population and clinic-
based studies 2018
USA,
Europe
Non-
Hispanic /
European 3617 cases HLA-B
1.22
(1.01-1.47)

(Wang et al)

8753
controls HLA-C
1.20
(1.02-1.41)

HLA class I
(increasing #)

8.00E-
04

HLA-DRB1
1.51
(1.27-1.78)

HLA-DQB1
1.30
(1.12-1.51)

HLA class II
(increasing #) <.0001
FL
International pooled
GWAS effort,
population and clinic-
based studies 2018
USA,
Europe
Non-
Hispanic /
European 2686 cases HLA-DRB1
1.54
(1.31-1.82)

(Wang et al)

8753
controls HLA-DQB1
1.42
(1.23-1.65)

HLA-DPB1
1.24
(1.10-1.40)

HLA class II
(increasing #) <.0001
CLL
Hematology clinic
patients and
population controls 2002 Germany
Non-
Hispanic /
European
45 female
cases
HLA-
DRB3/4/5 2.8 (NR) 0.01
33

(Mueller et al)

94 female
controls HLA-DQB1 4.4 (NR) 0.004

Transplant eligible
cases and population
donor based controls 2014 USA
Non-
Hispanic /
European

HLA-A
1.19
(1.07-1.33) 0.00016

(Gragert et al)

HLA-B
1.16
(1.02-1.32) 0.024

HLA-C
1.16
(1.02-1.31) 0.0077

International pooled
GWAS effort,
population and clinic-
based studies 2018
USA,
Europe
Non-
Hispanic /
European 2878 cases HLA-A
1.19
(1.02-1.38)

(Wang et al)

8753
controls HLA-DRB1
1.19
(1.00-1.42)

HLA-DQB1
1.20
(1.03-1.39)
MZL
International pooled
GWAS effort,
population and clinic-
based studies 2018
USA,
Europe
Non-
Hispanic /
European 741 cases HLA-B
1.34
(1.01-1.78)

(Wang et al)

8753
controls HLA-C
1.33
(1.04-1.70)

HLA class I
(increasing #)

0.026

HLA-DRB1
1.45
(1.12-1.89)

HLA class II
(increasing #) 0.0124

34

Table 2.2 HLA (and loci on chromosome 6p21.3) associations with DLBCL, FL, CLL, and MZL, based on imputation of SNP or genome-wide
association data. (Abbreviations: DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; CLL, chronic lymphocytic leukemia; OR, odds
ratio; RR, relative risk; NR, not reported)

Subtype
Study description
(reference) Year Country
Race /
ethnicity
# cases /
# controls SNP
Minor
Allele
Implicated
HLA
OR / RR
(95% CI) p-value
DLBCL
Pooled case-control
studies 2009
USA,
Canada,
Germany European 783 cases rs6457327 A C6orf15
0.78
(0.67-0.92) 7.00E-05

(Skibola et al)

3377 controls

Population-based case-
control studies, pooled
effort 2011 Denmark European 1592 cases rs10484561 G DQB1
1.36
(1.21-1.52) 1.41E-07

(Smedby et al)

6581 controls

Mayo Clinic case-control
study 2012 USA European 238 cases rs241447 G TAP2
1.38
(1.08-1.77) 1.10E-02

(Habermann et al)

1233 controls

International pooled
GWAS effort, population
and clinic-based studies 2012
USA,
Europe,
Canada,
Australia European 1881 cases rs3132453 A PRRC2A
0.70
(0.58-0.84) 1.89E-04

(Nieters et al)

7034 controls rs2239704 A TNF/LTA
0.78
(0.71-0.86) 8.98E-07

rs909253 G TNF/LTA
1.19
(1.08-1.31) 6.95E-04

rs2844482 T TNF/LTA
1.20
(1.08-1.33) 7.55E-04

Clinic-based patients and
community controls 2013 China Chinese 376 cases rs2647012 A DQB1
1.28
(1.02-1.61) 3.40E-02

(Qiao et al)

1542 controls

Hospital based patients
and community blood
donors 2013 Russia European 85 cases rs2647012 A DQB1
2.78
(1.47-5.25) 2.00E-03

(Weiner et al)

551 controls

International pooled
GWAS effort, population
and clinic-based studies 2014 USA European 3857 cases rs2523607 A B*08:01
1.32
(1.21-1.44) 2.40E-10

(Cerhan et al)

7666 controls

35

PAGE Consortium -
pooled cohort studies 2014 USA European 360 cases rs6457327 A C6orf15
0.81
(0.69-0.96) 1.30E-02

(Lim et al)

24183
controls

Hospital-based case
series and hospital
controls or population
controls 2015
Korea,
Hong
Kong,
Thailand
East
Asian 1124 cases rs2523607 A B*08:01
3.05
(1.32-7.05) 9.00E-03

(Bassig et al)

3596 controls

Danish lymphoma
registry and healthy
blood donors 2015 Denmark European 216 cases rs241447 G TAP2
0.61
(0.44-0.84) 3.00E-03

(Nielsen et al)

307 controls rs1800629 A TNF
1.46
(1.06-2.00) 1.90E-02

rs1799964 C TNF/LTA
0.48
(0.33-0.70) 1.00E-03
FL
U.S. population-cased
case-control study 2009
USA,
Canada,
Germany European 645 cases rs6457327 A C6orf15
0.59
(0.50-0.70) 4.70E-11

(Skibola et al)

3377 controls

International pooled
GWAS effort, population
and clinic-based studies 2010
USA,
Europe,
Canada,
Australia European 1465 cases rs10484561 G DQB1
1.95
(1.72-2.22) 1.12E-29

(Conde et al )

6011 controls rs7755224 G DQB1
2.07
(1.76-2.42) 2.00E-19

International pooled
GWAS effort, population
and clinic-based studies 2010
USA,
Europe,
Canada,
Australia European 1579 cases rs3132453 A PRRC2A
0.59
(0.48-0.73) 1.26E-06

(Nieters et al)

7034 controls

Meta-analysis,
population-based case-
control studies 2011
USA,
Canada,
Australia European 1428 cases rs2647012 A DQB1
0.70
(0.67-0.78) 4.00E-12

(Smedby et al)

6581 controls rs10484561 G DQB1
1.64 (1.45-
1.86) 5.00E-15

Clinic-based patients and
controls from the
Wellcome Trust Case- 2011 UK European 218 cases rs10484561 G DQB1
2.07
(1.61-2.63) 3.50E-09
36

Control Consortium 2

(Wrench D et al)

2691 controls rs6457327 A C6orf15
0.75
(0.61-0.93) 8.00E-03

Mayo Clinic-based case-
control study 2012 USA European 238 cases rs10484561 G DQB1
2.16
(1.66-2.81) 1.11E-88

(Cerhan et al)

1233 controls rs2647012 A DQB1
0.57
(0.46-0.70) 1.04E-07

rs241447 G TAP2
1.81
(1.46-2.24) 6.89E-08

Meta-analysis, population
based case-control
studies 2012
USA,
Scandina
via European 486 cases rs9275517 A DQA2
0.63
(0.55-0.73) 4.03E-11

(Skibola et al)

1472 controls rs3117222 A DPB1
0.66
(0.57-0.77) 1.45E-07

International pooled
GWAS effort, population
and clinic-based studies 2013
USA,
Israel European 2189 cases rs4530903 T DRB5/DQA1 1.93 2.69E-12

(Vijai et al)

6640 controls rs9268853 C DRB9 1.56 2.48E-10

rs2647046 A DQB1/DQA2 0.59 3.77E-10

rs2621416 G DQB2/TAP2 1.57 2.41E-09

PAGE Consortium -
pooled cohort studies 2014 USA European 318 cases rs6457327 A C6orf15
0.78
(0.64-0.95) 1.30E-02

(Lim et al)

24183
controls

International pooled
GWAS effort, population
and clinic-based studies 2014 USA European 4523 cases rs12195582 T DRA
1.78
(1.69-1.88) 5.36E-100

(Skibola et al)

Europe

13344
controls

DRB1*01
1.85
(1.70-2.03) 2.57E-42

DRB1*07:01
1.52
(1.39-1.66) 1.59E-20

rs17203612 T DQB1
1.44
(1.32-1.57) 4.59E-16

rs3130437 A C
1.23
(1.15-1.32) 8.23E-09

Danish lymphoma
registry and healthy
blood donors 2015 Denmark European 139 cases rs1799964 C TNF
0.51
(0.32-0.79) 3.00E-03

(Nielsen et al)

307 controls

37

Hospital-based cases
and healthy controls 2017
Malaysia,
Chinese Asian 83 cases rs2647012 A DQB1
0.56
(0.33-0.94) 2.90E-02
(Ten et al) 578 controls rs10484561 G DQB1
2.38
(1.31-4.34) 4.40E-03
CLL
Mayo Clinic-based case-
control study 2011 USA European 354 cases rs674313 T DRB5/DQA1
1.87
(1.47-2.38) 1.98E-07

(Slager et al)

1261 controls rs602875 A DRB5/DQA1 NR 8.10E-07

Meta analysis 2012 USA European 1196 cases rs210142 T BAK1
0.73
(0.68-0.79) 2.28E-16

(Slager et al)

2410 controls rs210134 A BAK1
0.75
(0.69-0.80) 8.34E-14

International pooled
GWAS effort, population
and clinic-based studies 2013
USA,
Europe European 3492 cases rs9273363 A DQB1 1.24 2.24E-10

(Berndt et al)

12228
controls

International CLL linkage
consortium and
Leukaemia Research
CLL4 trial patients and
controls from the
Wellcome Trust Case
Control Consortium 2013 England European 517 cases rs6904029 A A*02:01
1.32
(1.13-1.53) 1.38E-04
(Di Bernardo et al) 2930 controls
MZL
International pooled
GWAS effort, population
and clinic-based studies 2015
USA,
Europe European 1346 cases rs2922994 G B*08:01
1.64
(1.39-1.92) 2.43E-09
(Vijai et al) 3760 controls rs9461741 C BTNL2
2.66
(2.08-3.39) 3.95E-15

38

Table 2.3 Distribution of clinical stage for evaluated hematopoietic malignancies indicated for
transplant at the City of Hope at diagnosis

COH (BMT-indicated)

Clinical Stage

1 2 3 4
NHL 78
9.8% 107 13.6% 188 23.8% 417 52.8%

DLBCL 30
13.5% 46 20.6% 60 26.8% 88 39.1%

FL 9
5.6% 18 10.8% 47 28.0% 94 55.6%

Mantle 2
1.7% 4 3.4% 14 12.0% 97 82.9%

COH (Non-Hispanic Whites, overall, 1992-
2015)

Clinical Stage

1 2 3 4

NHL 575 17.6% 524 16.0% 741 22.7% 1,429 43.7%

DLBCL 300 20.2% 287 19.3% 370 24.9% 529 35.6%

FL 91 12.1% 93 12.4% 212 28.3% 353 47.1%

Mantle 26 6.8% 22 5.8% 49 12.9% 283 74.5%

SEER (Non-Hispanic Whites, Los Angeles, 1992-2013)

Clinical Stage

1 2 3 4

NHL 3,727 27.4% 2,146 15.8% 2,086 15.3% 5,634 41.4%

DLBCL 1,689 29.1% 1,150 19.8% 877 15.1% 2,091 36.0%

FL 850 29.7% 450 15.7% 647 22.6% 918 32.0%

Mantle 66 11.1% 58 9.8% 88 14.9% 380 64.2%

SEER 13 (Non-Hispanic Whites, 1992-2013)

Clinical Stage

1 2 3 4

NHL 27,151 30.3% 14,133 15.8% 13,887 15.5% 34,402 38.4%

DLBCL 12,239 31.8% 7,437 19.3% 5,747 14.9% 13,096 34.0%

FL 6,694 31.8% 3,292 15.7% 4,698 22.3% 6,344 30.2%

Mantle 468 11.3% 369 8.9% 601 14.5% 2,710 65.3%
Excludes unknown/missing

39

Table 2.4 Distribution of transplant status for evaluated NHL malignancies (DLBCL, FL, and MCL) in
Caucasian patients indicated for transplant at the City of Hope

Autologous Allogeneic Not Transplanted Total

DLBCL 113 49.3% 59 25.8% 57 24.9% 229
FL 79 41.4% 58 30.4% 54 28.3% 191
MCL 58 48.3% 33 27.5% 29 24.2% 120

40

Table 2.5 Frequencies of HLA Alleles in City of Hope *Imputed Donors Compared to National Marrow
Donor Program

COH NMDP
A
N = 6801 N = 1,242,890
A*01:01 12.18% 16.46%
A*02:01 28.91% 27.55%
A*02:05 1.36% 0.97%
A*03:01 12.27% 13.99%
A*11:01 5.81% 6.09%
A*23:01 2.01% 1.97%
A*24:02 10.56% 8.46%
A*25:01 1.37% 2.10%
A*26:01 3.82% 3.09%
A*29:02 2.89% 3.53%
A*30:01 1.51% 1.30%
A*30:02 1.08% 0.90%
A*31:01 3.20% 2.70%
A*32:01 2.84% 3.55%
A*33:01 1.15% 0.81%
A*68:01 3.49% 3.19%
A*68:02 1.54% 0.84%

B

B*07:02 11.20% 13.06%
B*08:01 7.33% 11.44%
B*13:02 2.33% 2.39%
B*14:02 4.27% 2.86%
B*15:01 5.60% 6.06%
B*18:01 4.50% 4.43%
B*27:05 2.53% 3.73%
B*35:01 6.86% 5.60%
B*35:02 2.28% 1.02%
B*35:03 2.42% 1.62%
B*38:01 25.90% 2.09%
B*39:01 1.27% 1.12%
B*39:06 1.04% 0.64%
B*40:01 4.75% 5.28%
B*40:02 1.49% 1.26%
B*44:02 8.21% 9.52%
B*44:03 4.45% 4.67%
B*49:01 1.83% 1.58%
41

B*50:01 1.30% 1.05%
B*51:01 6.54% 4.73%
B*52:01 1.68% 0.89%
B*55:01 1.41% 1.86%
B*57:01 2.47% 3.65%

C

C*01:02 3.53% 3.41%
C*02:02 3.80% 4.35%
C*03:03 4.51% 5.34%
C*03:04 7.57% 7.49%
C*04:01 14.40% 10.59%
C*05:01 8.08% 9.39%
C*06:02 7.57% 9.32%
C*07:01 12.40% 16.00%
C*07:02 13.40% 14.13%
C*07:04 1.37% 1.54%
C*08:02 5.03% 3.85%
C*12:02 1.51% 0.87%
C*12:03 5.39% 4.86%
C*14:02 1.54% 1.27%
C*15:02 2.89% 2.23%
C*16:01 3.08% 3.38%
C*17:01 1.26% 0.88%

DRB

DRB1*01:01 7.19% 8.60%
DRB1*01:02 2.78% 1.38%
DRB1*03:01 8.91% 12.16%
DRB1*04:01 7.81% 8.78%
DRB1*04:02 1.60% 1.06%
DRB1*04:03 1.40% 0.79%
DRB1*04:04 4.14% 3.88%
DRB1*04:07 2.12% 1.11%
DRB1*07:01 11.56% 13.42%
DRB1*08:01 3.03% 2.32%
DRB1*09:01 1.14% 1.03%
DRB1*11:01 5.96% 5.56%
DRB1*11:04 4.51% 2.95%
DRB1*12:01 1.75% 1.64%
DRB1*13:01 5.78% 5.63%
DRB1*13:02 4.04% 4.88%
DRB1*13:03 1.12% 1.09%
42

DRB1*14:01 3.04% 2.61%
DRB1*15:01 11.90% 13.46%
DRB1*15:02 1.40% 0.72%
DRB1*16:01 1.56% 1.43%

DQB

N=268
DQB1*02:01 18.30% 12.50%
DQB1*03:01 21.26% 19.80%
DQB1*03:02 12.20% 10.80%
DQB1*03:03 3.42% 3.50%
DQB1*04:02 4.40% 2.60%
DQB1*05:01 11.45% 10.30%
DQB1*05:02 1.84% 2.10%
DQB1*05:03 3.08% 2.60%
DQB1*06:01 1.37% 0.90%
DQB1*06:02 11.81% 11.90%
DQB1*06:03 6.12% 7.60%
DQB1*06:04 3.34% 3.40%

43

Table 2.6 Distribution of hematopoietic malignancies indicated for bone marrow transplant at the City
of Hope 1995-2015 and potential donors (controls without cancer) for whom high resolution HLA
haplotyping was imputed

Caucasian Hispanic Asian African American
N
Age
(median) N
Age
(median) N
Age
(median) N Age (median)
B-NHL 938 49 236 46 138 50 54 46
DLBCL 229 50 68 47 45 52 12 33
FL 191 47 49 46 18 43 5 46
CLL/SLL 124 51 12 53 4 45 9 41
Mantle 120 54 25 57 20 56 8 55
Other 274 48 82 49 51 49 20 45

Gender N % N % N % N %
Male 599 63.9% 153 64.8% 78 57.4% 38 70.4%
Female 339 36.1% 83 35.2% 58 42.6% 16 29.6%

N
Age
(median) N
Age
(median) N
Age
(median) N Age (median)
Donors 6801 40 2076 39 1029 39 365 42

44

Table 2.7 HLA Haplotype association with B-cell NHL subtypes (DLBCL, FL, and MCL), in Caucasian
patients and donors (analyses adjusted for sex and age)

Patient
(%)
Donor
(%) OR 95% CI p-value
DLBCL

n=229 n=6801

A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01

7.2% 7.2% 1.14 (0.67-1.93) 0.63
A*03:01~C*07:02~B*07:02~DRB1*15:01~DQB1*06:02

4.1% 4.3% 1.01 (0.5-2.05) 0.86
A*29:02~C*16:01~B*44:03~DRB1*07:01~DQB1*02:01

3.2% 2.4% 1.48 (0.67-3.29) 0.36
A*30:01~C*06:02~B*13:02~DRB1*07:01~DQB1*02:01

2.1% 1.6% 1.38 (0.46-4.1) 0.29
A*03:01~C*04:01~B*35:01~DRB1*01:01~DQB1*05:01

3.7% 1.5% 2.70 (1.21-6.01) 0.04
A*02:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01

3.1% 1.1% 3.27 (1.46-7.32) 0.016
A*26:01~C*12:03~B*38:01~DRB1*04:02~DQB1*03:02

3.0% 1.1% 3.06 (1.34-6.97) 0.01

A*01:01

24.3% 23.5% 1.11 (0.81-1.52) 0.5
A*02:01

46.0% 49.0% 0.94 (0.72-1.24) 0.7
A*03:01

22.6% 22.8% 0.99 (0.72-1.37) 1
A*26:01

9.7% 7.4% 1.30 (0.83-2.06) 0.3
A*29:02

6.6% 5.7% 1.20 (0.7-2.08) 0.5
A*30:01

3.1% 2.9% 1.11 (0.51-2.42) 0.8

B*07:02

17.5% 21.7% 0.79 (0.56-1.12) 0.2
B*08:01

15.0% 15.2% 1.06 (0.73-1.55) 0.8
B*13:02

8.3% 4.8% 1.99 (1.21-3.26) 0.007
B*35:01

10.5% 11.9% 0.89 (0.57-1.37) 0.6
B*38:01

8.3% 5.1% 1.52 (0.92-2.49) 0.1
B*44:03

9.6% 8.7% 1.13 (0.72-1.79) 0.6

C*04:01

22.9% 22.1% 0.76 (0.54-1.06) 0.1
C*06:02

23.5% 14.9% 1.72 (1.24-2.4) 0.001
C*07:01

17.4% 23.2% 0.71 (0.49-1.02) 0.06
C*07:02

19.7% 24.5% 0.80 (0.57-1.14) 0.2
C*12:03

13.6% 10.4% 1.26 (0.84-1.9) 0.3
C*16:01

7.5% 5.8% 1.32 (0.77-2.24) 0.3

DRB1*01:01

16.0% 13.5% 1.20 (0.82-1.75) 0.34
DRB1*03:01

16.0% 17.9% 0.98 (0.67-1.42) 0.9
DRB1*04:02

4.6% 3.0% 1.40 (0.72-2.74) 0.3
DRB1*07:01

23.3% 22.9% 1.03 (0.75-1.43) 0.9
DRB1*15:01

17.8% 23.0% 0.70 (0.49-1) 0.05

45

DQB1*02:01

34.4% 35.0% 1.03 (0.76-1.4) 0.9
DQB1*03:02

16.7% 21.1% 0.82 (0.56-1.21) 0.3
DQB1*05:01

26.6% 22.0% 1.19 (0.86-1.66) 0.3
DQB1*06:02

17.7% 22.2% 0.74 (0.51-1.09) 0.1

Patient
(%)
Donor
(%) OR 95% CI p-value
FL

n=191 n=6801

A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01

7.2% 7.2% 1.13 (0.64-1.97) 0.68
A*29:02~C*16:01~B*44:03~DRB1*07:01~DQB1*02:01

4.4% 2.4% 2.01 (0.96-4.22) 0.1
A*23:01~C*04:01~B*44:03~DRB1*07:01~DQB1*02:01

3.0% 1.1% 2.69 (0.95-7.58) 0.17
A*02:01~C*06:02~B*13:02~DRB1*07:01~DQB1*02:01

2.8% 1.1% 2.70 (1.09-6.66) 0.04

A*01:01

28.0% 23.5% 1.33 (0.96-1.85) 0.09
A*02:01

48.4% 49.0% 1.02 (0.76-1.37) 0.9
A*23:01

7.7% 3.9% 1.90 (1.07-3.34) 0.03
A*29:02

10.4% 5.7% 1.97 (1.2-3.22) 0.007

B*08:01

17.6% 15.2% 1.29 (0.86-1.93) 0.2
B*13:02

9.4% 4.8% 2.21 (1.30-3.77) 0.004
B*44:03

14.1% 8.7% 1.77 (1.13-2.76) 0.01

C*04:01

26.8% 26.9% 0.98 (0.68-1.41) 0.9
C*06:02

24.8% 14.9% 1.86 (1.27-2.71) 0.001
C*16:01

9.8% 5.8% 1.81 (1.04-3.14) 0.04

DRB1*03:01

16.8% 17.9% 1.02 (0.67-1.54) 0.9
DRB1*07:01

41.3% 22.9% 2.41 (1.75-3.31) <.0001

DQB1*02:01

45.3% 35.0% 1.63 (1.16-2.3) 0.005

Patient
(%)
Donor
(%) OR 95% CI p-value
MCL

n=120 n=6801

A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01

8.2% 7.2% 1.42 (0.72-2.80) 0.37
A*03:01~C*07:02~B*07:02~DRB1*15:01~DQB1*06:02

4.8% 4.3% 1.10 (0.43-2.79) 0.7
A*02:01~C*03:04~B*15:01~DRB1*04:01~DQB1*03:02

2.1% 2.0% 1.20 (0.30-4.81) 0.8
A*26:01~C*12:03~B*38:01~DRB1*04:02~DQB1*03:02

3.0% 1.1% 3.13 (1.00-9.83) 0.06

46

A*01:01

30.6% 23.5% 1.60 (1.05-2.43) 0.03
A*02:01

46.8% 49.0% 1.03 (0.7-1.51) 0.9
A*03:01

18.0% 22.8% 0.69 (0.42-1.14) 0.2
A*26:01

9.0% 7.4% 1.24 (0.63-2.44) 0.5

B*07:02

12.5% 21.7% 0.98 (0.61-1.59) 0.9
B*08:01

17.8% 15.2% 1.28 (0.76-2.14) 0.4
B*15:01

11.2% 10.6% 1.14 (0.61-2.11) 0.7
B*38:01

6.5% 5.1% 1.04 (0.46-2.34) 0.9

C*03:04

15.0% 12.5% 1.46 (0.83-2.58) 0.19
C*07:01

33.0% 23.2% 1.65 (1.07-2.54) 0.02
C*07:02

19.0% 24.5% 0.76 (0.45-1.27) 0.3
C*12:03

10.0% 10.4% 0.84 (0.43-1.66) 0.6

DRB1*03:01

25.0% 17.9% 1.68 (1.05-2.67) 0.03
DRB1*04:01

17.3% 14.3% 1.47 (0.86-2.49) 0.2
DRB1*04:02

5.8% 3.0% 1.82 (0.76-4.33) 0.2
DRB1*08:01

6.7% 4.3% 1.43 (0.64-3.18) 0.4

DQB1*02:01

37.4% 35.0% 1.15 (0.74-1.78) 0.5
DQB1*03:02

20.9% 21.1% 1.25 (0.74-2.12) 0.4
DQB1*06:02 15.4% 22.2% 0.57 (0.32-1.03) 0.06

47

Table 2.8 HLA Haplotype association with B-cell NHL subtypes (DLBCL and FL), in Hispanic patients and
donors (analyses adjusted for sex and age)

Patient
(%)
Donor
(%) OR 95% CI p-value
DLBCL

n=68 n=2026

A*24:02~C*07:02~B*39:06~DRB1*14:06~DQB1*03:01

1.5% 3.0% 0.49 (0.07-3.58) 0.5
A*02:01~C*04:01~B*35:12~DRB1*08:02~DQB1*04:02

3.5% 2.8% 1.35 (0.35-5.18) 0.6
A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01

1.5% 2.4% 0.67 (0.09-4.95) 0.7
A*30:01~C*06:02~B*13:02~DRB1*07:01~DQB1*02:01

3.8% 1.7% 2.23 (0.56-8.85) 0.3

A*01:01

8.3% 9.9% 0.80 (0.31-2.03) 0.6
A*02:01

61.7% 42.2% 2.19 (1.28-3.73) 0.004
A*24:02

21.7% 30.3% 0.64 (0.34-1.19) 0.2
A*30:01

8.3% 3.5% 2.42 (0.92-6.37) 0.07

B*08:01

5.2% 6.0% 0.86 (0.26-2.83) 0.8
B*13:02

8.6% 2.8% 3.06 (1.14-8.21) 0.03
B*35:12

10.3% 7.0% 1.63 (0.68-3.93) 0.3
B*39:06

1.7% 7.7% 0.19 (0.03-1.39) 0.1

C*04:01

32.7% 36.7% 0.79 (0.44-1.44) 0.4
C*06:02

13.5% 7.9% 1.65 (0.72-3.80) 0.2
C*07:01

11.5% 12.7% 0.99 (0.41-2.36) 1
C*07:02

23.1% 31.1% 0.73 (0.38-1.42) 0.4

DRB1*03:01

9.8% 11.3% 0.84 (0.35-1.98) 0.7
DRB1*07:01

24.6% 13.6% 2.01 (1.09-3.70) 0.3
DRB1*08:02

14.8% 9.1% 1.68 (0.81-3.53) 0.2
DRB1*14:06

3.3% 10.9% 0.25 (0.06-1.05) 0.06

DQB1*02:01

36.4% 23.8% 1.73 (0.98-3.06) 0.6
DQB1*03:01

27.3% 31.6% 0.77 (0.42-1.42) 0.4
DQB1*04:02

21.8% 29.5% 0.67 (0.35-1.29) 0.2

Patient
(%)
Donor
(%) OR 95% CI p-value
FL

n=49 n=2026

A*24:02~C*07:02~B*39:06~DRB1*14:06~DQB1*03:01

2.0% 3.0% 0.82 (0.11-6.11) 0.7
A*68:02~C*08:02~B*14:02~DRB1*01:02~DQB1*05:01

6.9% 2.3% 3.23 (0.99-10.52) 0.07
A*03:01~C*04:01~B*35:01~DRB1*01:01~DQB1*05:01

11.5% 1.2% 8.71 (2.06-36.88) 0.0033
48

A*03:01

18.8% 9.6% 2.31 (1.08-4.89) 0.03
A*24:02

20.8% 30.3% 0.60 (0.30-1.22) 0.2
A*68:02

6.3% 4.4% 1.51 (0.45-5.01) 0.5
B*14:02

10.6% 11.3% 0.96 (0.37-2.48) 0.9
B*35:01

21.3% 16.3% 1.34 (0.65-2.74) 0.4
B*39:06

6.4% 7.7% 0.78 (0.24-2.56) 0.7

C*04:01

39.0% 36.7% 1.05 (0.55-2.00) 0.9
C*07:02

19.5% 31.1% 0.60 (0.27-1.31) 0.2
C*08:02

14.6% 12.2% 1.29 (0.53-3.12) 0.6

DRB1*01:01

27.3% 6.9% 4.83 (2.38-9.78) <0.001
DRB1*01:02

9.1% 9.1% 1.01 (0.35-2.89) 1
DRB1*14:06

11.4% 10.9% 1.01 (0.39-2.63) 1

DQB1*03:01

42.1% 31.6% 1.51 (0.78-2.93) 0.2
DQB1*05:01

39.5% 19.4% 2.65 (1.35-5.79) 0.0045

49

Table 2.9 HLA Haplotype association with DLBCL, in Asian patients and donors (analyses adjusted for
sex and age)

Patient
(%)
Donor
(%) OR 95% CI p-value
DLBCL

n=45 n=1029

A*33:03~C*03:02~B*58:01~DRB1*03:01~DQB1*02:01

3.1% 5.9% 0.59 (0.10-3.51) 0.6
A*11:01~C*08:01~B*15:02~DRB1*12:02~DQB1*03:01

4.9% 3.9% 1.09 (0.26-4.54) 0.7
A*34:01~C*07:02~B*38:02~DRB1*15:02~DQB1*05:02

4.4% 1.3% 4.28 (0.91-20.1) 0.07

A*11:01

30.0% 35.3% 0.71 (0.35-1.44) 0.34
A*33:03

7.5% 15.2% 0.47 (0.14-1.58) 0.22
A*34:01

14.5% 7.3% 2.76 (1.14-6.69) 0.024

B*15:02

15.0% 10.1% 1.41 (0.57-3.51) 0.46
B*38:02

15.0% 9.3% 1.95 (0.78-4.86) 0.15
B*58:01

7.5% 9.8% 0.80 (0.24-2.69) 0.72

C*03:02

4.9% 10.6% 0.44 (0.10-1.87) 0.27
C*07:02

26.8% 30.2% 0.87 (0.43-1.77) 0.69
C*08:01

24.4% 17.1% 1.53 (0.72-3.21) 0.27

DRB1*03:01

8.1% 8.4% 1.16 (0.34-3.95) 0.81
DRB1*12:02

32.4% 19.2% 2.18 (1.05-4.49) 0.036
DRB1*15:02

21.6% 21.9% 1.04 (0.46-2.34) 0.93

DQB1*02:01
18.5% 16.0% 1.34 (0.49-3.71) 0.57
DQB1*03:01

51.9% 37.6% 1.58 (0.72-3.49) 0.26
DQB1*05:02

14.8% 15.6% 1.31 (0.43-3.97) 0.64

50

Table 2.10 HLA Haplotype association with CLL/SLL subtypes, in Caucasian patients and donors
(analyses adjusted for sex and age)

Patient
(%)
Donor
(%) OR 95% CI p-value
CLL/SLL

n=124 n=6801

A*01:01~C*07:01~B*08:01~DRB1*03:01~DQB1*02:01

6.7% 7.2% 1.09 (0.55-2.16) 0.8
A*03:01~C*07:02~B*07:02~DRB1*15:01~DQB1*06:02

4.5% 4.3% 1.10 (0.47-2.57) 0.7
A*02:01~C*05:01~B*44:02~DRB1*04:01~DQB1*03:01

4.5% 2.7% 1.64 (0.63-4.32) 0.3
A*29:02~C*16:01~B*44:03~DRB1*07:01~DQB1*02:01

2.9% 2.4% 1.35 (0.48-3.74) 0.6

A*01:01

27.6% 23.5% 1.33 (0.88-2.02) 0.2
A*02:01

52.6% 49.0% 1.27 (0.88-1.85) 0.2
A*03:01

25.0% 22.8% 1.11 (0.72-1.71) 0.6
A*29:02

10.3% 5.7% 1.85 (1.00-3.45) 0.05

B*07:02

19.5% 21.7% 0.89 (0.55-1.43) 0.6
B*08:01

17.7% 15.2% 1.29 (0.78-2.11) 0.3
B*44:02

17.7% 15.7% 1.23 (0.75-2.02) 0.4
B*44:03

9.7% 8.7% 1.10 (0.58-2.09) 0.8

C*05:01

18.1% 15.5% 1.25 (0.75-2.08) 0.4
C*07:01

29.5% 23.2% 1.44 (0.93-2.21) 0.1
C*07:02

24.8% 24.5% 1.07 (0.68-1.69) 0.8
C*16:01

6.7% 5.8% 1.10 (0.50-2.43) 0.8

DRB1*03:01

18.5% 17.9% 1.14 (0.69-1.88) 0.6
DRB1*04:01

22.2% 14.3% 1.88 (1.18-3.01) 0.008
DRB1*07:01

21.3% 22.9% 0.91 (0.57-1.46) 0.7
DRB1*15:01

15.7% 23.0% 0.59 (0.35-1.00) 0.05

DQB1*02:01

33.3% 35.0% 0.96 (0.62-1.48) 0.8
DQB1*03:01

36.5% 38.4% 0.89 (0.58-1.36) 0.6
DQB1*06:02

14.6% 22.2% 0.58 (0.32-1.03) 0.06

51

Table 2.11 Effect of homozygosity at the three HLA class I loci -A, -B and -C and three HLA class II loci -DRB1, DQB1, and DPB1 on susceptibility
to three B-cell NHL subtypes (DLBCL, FL, and MCL) in Caucasian patients and donors (analyses adjusted for sex and age)

Donors
(n=6801)
DLBCL
(n=229)
FL
(n=191)
MCL
(n=120)
% % OR (95% CI) % OR (95% CI) % OR (95% CI)
Class I locus

HLA-A Heterozygote 84.5% 82.5% 1.00 (REF)

86.4% 1.00 (REF)

81.7% 1.00 (REF)

Homozygote 15.5% 17.5% 1.18 (0.83-1.68)

13.6% 0.87 (0.57-1.33)

18.3% 1.26 (0.78-2.04)
HLA-B Heterozygote 93.7% 92.1% 1.00 (REF)

90.6% 1.00 (REF)

90.8% 1.00 (REF)

Homozygote 6.3% 7.9% 1.27 (0.78-2.10)

9.4% 1.54 (0.93-2.53)

9.2% 1.43 (0.75-2.73)
HLA-C Heterozygote 89.8% 87.3% 1.00 (REF)

87.4% 1.00 (REF)

87.5% 1.00 (REF)

Homozygote 10.2% 12.7% 1.28 (0.86-1.92)

12.6% 1.26 (0.81-1.95)

12.5% 1.23 (0.70-2.15)

Total # of homozygote loci 0 75.5% 72.5% 1.00 (REF)

75.4% 1.00 (REF)

72.5% 1.00 (REF)

1 18.6% 20.5% 1.17 (0.84-1.63)

16.2% 0.89 (0.60-1.31)

18.3% 1.06 (0.65-1.71)

2+ 5.9% 7.0% 1.26 (0.75-2.15)

8.4% 1.43 (0.84-2.43)

9.2% 1.62 (0.84-3.11)
p-trend

0.84

0.18

0.5
Class II locus

HLA-DRB1 Heterozygote 91.9% 85.9% 1.00 (REF)

91.9% 1.00 (REF)

93.3% 1.00 (REF)

Homozygote 8.1% 14.1% 1.89 (1.28-2.79)

8.1% 2.06 (1.37-3.09)

6.7% 0.80 (0.38-1.66)
HLA-DQB1 Heterozygote 86.0% 76.1% 1.00 (REF)

86.0% 1.00 (REF)

85.0% 1.00 (REF)

Homozygote 14.0% 23.9% 1.46 (1.04-2.06)

14.0% 1.50 (1.04-2.16)

15.0% 1.09 (0.65-1.83)
HLA-DPB1* Heterozygote 75.8% 74.8% 1.00 (REF)

80.3% 1.00 (REF)

80.5% 1.00 (REF)

Homozygote 24.2% 25.2% 1.03 (0.65-1.62)

19.7% 0.76 (0.42-1.37)

19.5% 0.71 (0.31-1.61)

Total # of homozygote loci* 0 65.1% 62.6% 1.00 (REF)

73.2% 1.00 (REF)

70.7% 1.00 (REF)

1 26.3% 26.2% 1.06 (0.67-1.68)

15.5% 0.53 (0.27-1.02)

26.8% 0.97 (0.47-2.03)
2+ 8.6% 11.2% 1.60 (0.84-3.07)

11.3% 1.30 (0.61-2.79)

2.4% 0.33 (0.04-2.48)
p-trend 0.49 0.03 0.39
*Restricted to those with DPB molecular typing: 4594 Donors, 107 DLBCL, 74 FL, 41 MCL
52

Chapter 3. Artificial Light at Night at Non-Hodgkin Lymphoma Risk in the California Teachers Study
Cohort
3.1. Introduction
Artificial light at night (ALAN) is necessary for modern society, but modern electric lights and
other electronic devices elevate ambient light exposures over those experienced in the course of human
evolution up to the industrial era. Major contributors to outdoor lighting levels are commercial lighting,
industrial lighting, street lighting, and residential lights (Elvidge, Baugh, Zhizhin, Hsu, & Ghosh, 2017)
while indoor light exposure is driven by a combination of electric lights, electronics, and exposure to
outdoor lights through window when those are extinguished. Outdoor ALAN has steadily increased for
decades, at accelerating rates in recent years (Kyba et al., 2017). Outdoor ALAN has been a concern of
astronomers for sixty years, and the influence of light exposure on sleep has become a hot topic of
investigation. Epidemiological studies have linked satellite-measured outdoor lighting to adverse health
outcomes associated with circadian disruption. Whether this is a direct pathway (outdoor lighting
affects humans in their sleeping habitats) or a proxy measure for either indoor light exposure or other
correlated factors has not yet been conclusively settled. Growing evidence implicates exposure to ALAN
with circadian disruption in humans (Blask et al., 2014; Qian, Block, Colwell, & Matveyenko, 2013), but
the few studies that have been conducted comparing direct satellite measured ALAN and bedroom
exposure have found little to no correlation (Huss et al., 2019; Rea, Brons, & Figueiro, 2011). We
attempted to improve upon use of satellite measured ALAN through use of the New World Atlas of
Artificial Night Sky Brightness. This chapter investigates the association between ALAN exposure and
NHL, which are presumed to be associated through the mechanism of chronic inflammation.
It is recognized that sleep plays a role in immune function, and we are just beginning to
understand the underlying mechanisms (Irwin, 2015). Exposure to light at night suppresses melatonin
53

production and disrupts circadian rhythm (Lewy, Wehr, Goodwin, Newsome, & Markey, 1980; Reiter et
al., 2007). Melatonin reduces arousal and lowers core body temperature, thereby promoting sleep
(Blask, 2009). Melatonin has also been shown to modulate activation of the inflammatory nuclear
transcription factor kappa beta (NF-κβ) pathway, which modulates levels of several immune cytokines
such as interleukins (IL) -2, -6, -10, tumor necrosis factor (TNF), and C-reactive protein (CRP)(Favero,
Franceschetti, Bonomini, Rodella, & Rezzani, 2017). Disruptions of circadian rhythm are classified as a
group 2A carcinogen by IARC (probably carcinogenic), based largely on increased breast cancer risk
among night-shift workers (Straif et al., 2007). Satellite-measured ALAN exposure has also been
associated with cancer of several other sites including, lung (Al-Naggar & Anil, 2016), breast (Al-Naggar
& Anil, 2016; Hurley et al., 2014; James et al., 2017), colorectal (Al-Naggar & Anil, 2016), and prostate
(Al-Naggar & Anil, 2016; Kim, Lee, Kim, & Kim, 2017), and with laboratory models confirming the
mechanism of melatonin suppression as a facilitator of tumor growth (Blask et al., 2014; Giudice et al.,
2018). A growing body of evidence is suggesting that ALAN exposure contributes to poorer health
overall; including increased risk of other proinflammatory conditions such as diabetes(Koo et al., 2016).
A majority of these studies are conducted on an ecological scale, and therefore have little control for
confounding.
Studies have linked circulating levels of inflammatory markers to sleep deprivation, but the
effects of the specific markers vary by the amount and type of sleep deprivation. For example, some
studies show no effects on IL-6 or TNF over 4 days of sleep deprivation, but another reported elevation
in IL-6 with 10 days of deprivation (Haack, Sanchez, & Mullington, 2007; Meier-Ewert et al., 2004; Sauvet
et al., 2010; Shearer et al., 2001; van Leeuwen et al., 2009). These studies have largely been conducted
in very controlled environments among a relatively modest number of healthy volunteers, have
measured the effects of acute sleep deprivation (rather than long-term or chronic sleep deprivation),
and have measured only a few targeted immune markers. There is some evidence from epidemiologic
54

studies suggesting that self-reported measures of prolonged sleep deprivation, as well as poor sleep
quality are associated with markers of chronic inflammation, including CRP, TNF, IL-2, and IL-6(Irwin,
2015). Other studies have reported elevated CRP levels among those with shorter sleep duration in
women and not men (Liukkonen et al., 2007; Miller et al., 2009). Studies in mice have shown increased
circulating levels of proinflammatory cytokines when exposed to both PM
2.5
and light at night (Hogan,
Kovalycsik, Sun, Rajagopalan, & Nelson, 2015).
Several sources of chronic inflammation have been implicated in NHL etiology (Smedby &
Ponzoni, 2017) and several studies suggest a possible link between sleep disruption and NHL. A Finnish
study has previously associated shift work with increased risk of non-Hodgkin lymphoma (NHL) in men,
and while elevated, not significantly in women (Lahti, Partonen, Kyyrönen, Kauppinen, & Pukkala, 2008).
A study evaluating cancer risk based on time zones found increased risk of chronic lymphocytic leukemia
(CLL) among those living closer to the western edge of a time zone, based on the hypothesis that living
in a more western region of a time zone results in later onset of nighttime, relative to clock time. There
has also been reported evidence of epigenetic silencing of the circadian clock gene Cryptochrome 1
among CLL patients (Hanoun et al., 2012). One of the challenges in sleep research is disentangling the
independent effects of sleep from those of circadian disruption. An important step in this process is
understanding the degree to which inter-individual variability in circadian rhythms, or circadian
phenotypes, influence an individual’s vulnerability to the impact of sleep deficiency. Chronotype,
typically being defined as being a morning or evening person is the most widely-considered and
validated self-reported measure of circadian phenotype (Goel, Basner, Rao, & Dinges, 2013; Wulff,
Porcheret, Cussans, & Foster, 2009). To better understand the effect ALAN exposure may have on NHL
risk, we evaluated the risk in the California Teachers Study cohort.
55

3.2. Methods
3.2.1. Study Population
The California Teachers Study (CTS) is a large prospective cohort recruited from active or retired
teachers and members of the California State Teachers Retirement System in 1995. A total of 133,479
women returned the baseline questionnaire and have participated in ongoing follow-up activities.
Although originally designed to study breast cancer, the study has been expanded to look at several
other outcomes including cardiovascular disease, Parkinson’s, and other cancers, including NHL (E. T.
Chang et al., 2011; Ellen T. Chang et al., 2011; Gatto et al., 2014; Yani Lu et al., 2009; Ostro et al., 2015;
Ostro et al., 2010). ALAN has previously been associated with breast cancer risk in the CTS (HR: 1.12,
95% CI: 1.00, 1.26) using the coarser 2.7km data(Hurley et al., 2014). The CTS represents a broad age
range (22-104 years at baseline, median age 53) in both urban and rural areas. Complete details of the
cohort have been published previously (Bernstein, Allen, Anton-Culver, Deapen, & et al., 2002). The
baseline questionnaire captured detailed information on diet, alcohol use, height/weight, physical
activity, smoking, environmental exposures, medication use, and personal/family history of cancer
including NHL. The most recent follow-up questionnaire (questionnaire 5) was conducted in 2013-2014
with a response rate of 67%, and included questions about sleep quality. All questionnaires have been
digitally scanned, coded, and cleaned.
For our analysis, participants were excluded if they: lived outside of California at baseline or
could not be geocoded (n=14,427), had prevalent cancer (n=8,078), or did not answer/reported sleeping
at night with a light on in the bedroom (n=5,488), for a final analytic cohort of 105,486 participants
(Figure 3.1).
56

3.2.2. Exposure Assessment
Unlike the previous study in the CTS utilizing the US Defense Meteorological Satellite
Operational Linescan System (DMSP-OLS), and more recent studies that utilize the National
Aeronautic and Space Administration (NASA)/National Oceanic and Atmospheric Administration
(NOAA) VIIRS satellite; we will assess ALAN using the New World Atlas of Artificial Night Sky Brightness
(Falchi et al., 2016). The DMSP-OLS satellite has data available since the 1990s at a 2.7km spatial
resolution, providing longer historical data. However, limitations such as the coarse resolution and
oversaturation of urban areas led to the launching of the VIIRS satellite in 2011 (Baugh, Elvidge,
Tilottama, & Ziskin, 2010; Xuecao & Zhou, 2017). The VIIRS satellite has provided improved
resolution (750m) and better dynamic range (Elvidge et al., 2017). The New World Atlas consists of
thousands of observations using handheld sky quality meters and calibrated using VIIRS data to create a
global measure of ALAN. Unlike using satellite data alone, the New World Atlas better captures the
exposure on the ground (Simons, Yin, and Longcore, unpublished data). This method produces a global
measure of luminance at the zenith in mcd/m
2
. The cohort had a median exposure of 2.4 mcd/m
2
(IQR
1.22-3.95, Table 3.3). For reference, a natural, moonless night sky would only have a brightness of 0.2
mcd/m
2
and by 2 mcd/m
2
the Milky Way would no longer be visible (Bortle, 2001). The VIIRS data that
are used to create the New World Atlas measure are based on 2015 annual average data; within satellite
comparisons of DMSP-OLS and VIIRS show high (>0.90) correlation between years, and only modest
correlation (0.70) between the two satellites (Table 3.1). We also aggregated the VIIRS data to a 2.7km
resolution to compare with the DMSP-OLS; and had good correlation with the regular VIIRS (0.89) and
the DMSP-OLS (0.80). Due to the non-normal distribution of light at night, we categorized our exposure
by quintiles (Figure 3.2).
CTS participant’s address at baseline have been previously geocoded. ALAN was assigned to
each participant’s address at baseline in 1995-1996. We expect a level of misclassification due to the
57

recent availability of the data; however, the major source of ALAN is street lighting, which is stable
over 15-30 year periods (Kyba & Aronson, 2015; Shaflik, 1997). Of the 113,441 geocoded addresses,
there were duplicates for 7,473 participants (Table 3.2). A majority of these duplicates appeared to
be in smaller towns. We conducted sensitivity analyses by jittering geocoded addresses, with a
radius of 1km. There was extremely high correlation between the original and jittered values
(0.996, Table 3.1) therefore the original values were in the analysis. All light at night measures were
assigned using the raster package in R.
3.2.3. Outcome Assessment
Our cases were obtained through linkage to the California Cancer Registry. Non-Hodgkin
lymphoma (NHL) was defined as having a diagnosis with the International Classification of Diseases for
Oncology, third edition codes of: diffuse large B-cell lymphoma (DLBCL: 9678, 9679, 9680, 9684),
follicular lymphoma (FL: 9690, 9691, 9695, 9698), mantle cell lymphoma (MCL: 9673), marginal zone
lymphoma (MZL: 9699), chronic lymphocytic leukemia/Small lymphocytic lymphoma (CLL/SLL: 9670,
9823), Burkitt lymphoma (BL: ICD-O-3 codes 9687, 9826), and other B cell lymphomas (other, including
not otherwise specified (NOS): 9590, 9591, 9596, 9671, 9675, 9727, 9728, 9833, 9835, 9836, 9761). We
identified 933 incident cases of NHL in the cohort from baseline through December 31, 2016.
3.2.4. Covariates
We assessed confounding among several covariates including; age, race (non-Hispanic White,
Black, Hispanic, Asian/Pacific Islander, Other), alcohol and smoking status (current, former, never), body
mass index (<18.5, 18.5-25, 25-30, or ≥30 kg/m
2
), United States Census block level socioeconomic status
(statewide quartiles), United States Census block urban/rural status (Ratcliffe, Burd, Holder, & Fields,
2016), ultraviolet radiation, use of non-steroidal anti-inflammatory drugs, exercise (meeting/not
58

meeting American Heart Association recommended physical activity levels (Troiano, 2018)), exposure to
pesticides, hormone use, and family history of hematologic malignancy.
3.2.5. Statistical Methods
We evaluated the association between ALAN and NHL overall and among the three most
common subtypes (DLBCL, FL, CLL/SLL). We calculated hazard ratios (HR) and 95% confidence intervals
(95% CI) using a Cox proportional hazards model. In the Cox regression models, the time scale (in days)
was defined by age at entry into the cohort and the first of the following ages: at event (NHL diagnosis),
at censoring (e.g., when a participant moved out of California for more than four months), at death, or
at end of follow-up (December 31, 2016). Race was the only covariate that altered HR estimates by 10%
or greater and was included in the final model. Additionally, the association between self-reported
sleep and ALAN was assessed through logistic regression among participants who responded to
questionnaire 5. HR and 95% CI were calculated for the 91 cases of NHL diagnosed between
questionnaire 5 and end of follow-up (December 31, 2016). All analyses were conducted in SAS 9.4 (SAS
Institute Inc., Cary, NC).
3.3. Results
3.3.1. Main Effects of Artificial Light at Night
The majority of the cohort (87%) was non-Hispanic White (Table 3.3). Almost half the cohort
(45%) resided in the highest SES quartile census block as defined by the US Census while only 14%
resided in a rural census block. The most common NHL diagnosis was DLBCL (n=234), followed by
CLL/SLL (n=229) and FL (n=179). The largest proportion of the cohort resided in Los Angeles County
(n=52,370), followed by Orange County (n=23,082) and San Diego County (n=21,728) (Figure 3.3). These
counties also had the most NHL cases (Figure 3.5). The median age at study entry by county ranged
from 46-58 years old (Figure 3.4). The highest levels of ALAN were in Los Angeles County with a median
59

of 5.28 mcd/m
2
and an interquartile range of 3.57-6.93 (Figure 3.6). As expected, many of the northern
counties had the lowest levels of ALAN.
Compared to participants in the lowest quintile of ALAN, those in the highest quintile had 30%
increased risk (95% CI 1.06-1.59) of developing NHL (Table 3.4). Covariates that were associated with
NHL risk included higher SES (HR 1.81, 95% CI 1.18-2.78) and family history of hematologic malignancy
(HR 1.49, 95% CI 1.17-1.90), while current drinkers at baseline were associated with decreased NHL risk
(HR 0.73, 95% CI 0.55-0.95). After adjusting for age and race, the highest quintile of ALAN was
associated with a 35% increased risk of NHL (95% CI 1.10-1.66).
When stratified by subtype, the association with ALAN was only apparent for DLBCL (HR 1.74,
95% CI 1.13-2.70). Additional heterogeneity was seen in the highest quintile of UV exposure only being
protective for CLL/SLL (HR 0.56, 95% CI 0.33-0.94). While higher BMI was trending towards increased
risk for DLBCL and FL, it appeared to be protective for CLL/SLL. The age and race adjusted risk of DLBCL
was elevated in all quintiles of ALAN, and significantly, 81% greater in the highest quintile of ALAN (95%
CI 1.17-2.82). Additional inclusion of physical activity in the model further strengthen the association
between ALAN and NHL (HR 1.44 95% CI 1.15-1.81, Table 3.4).
3.3.2. Sensitivity Analyses
The association between NHL and ALAN did not change when using the jittered World Atlas
values. When looking at the other satellite measures (Table 3.5), the highest quintile of both the VIIRS
(HR 1.08, 95% CI 0.87-1.33) and 2.7km aggregated VIIRS (HR 1.11, 95% CI 0.90-1.37) only showed
modest, non-significantly increased risk. The highest quintile of the DMSP-OLS was also marginally not
statistically significantly associated with increased NHL risk (HR 1.21, 95% CI 0.98-1.49). While exercise
was not significantly associated with NHL (HR 1.05, 95% CI 0.91-1.20), when tested against the age
adjusted model, it appeared to be a confounder (HR 1.45, 95% CI 1.15-1.82). Additional inclusion of BMI
60

and SES further attenuated the effect of ALAN (HR 1.32, 95% CI 1.06-1.64). Compared to participants in
the interquartile range of ALAN, those in the top 5% were at increased risk of developing NHL (HR 1.39,
95% CI 1.04-1.85). There was no association between sleeping with a light on at night and NHL (HR 0.97,
95% CI 0.71-1.34).
3.3.3. Questionnaire 5 (2012-2013)
Of the 105,486 eligible participants, 56,114 responded to questions about sleep history in
questionnaire 5 (Figure 3.7). Table 5 contains the 40,613 participants that reported the same sleep
habits over the past year. The mean age of the participants who replied was 67.2 years old and a
majority was non-Hispanic White (89%). A majority reported having a morning chronotype (59%) and
most reported either very good or fairly good sleep quality (86%). Only 32% reported sleeping the
recommended 8 hours or more, with the majority (42%) reporting sleeping for 7 hours. Participants
living in the highest quintile of ALAN were 1.16 times as likely (95% CI 1.06-1.28, Table 3.6) to report
poor sleep quality than very good sleep quality, adjusted for age, race, and chronotype. In addition,
compared to the lowest quintile of ALAN, the highest quintile was 1.22 (95% CI 1.15-1.30) times as likely
to report sleeping less than 8 hours a night. For the 91 NHL were diagnosed with follow-up through
December 31, 2016 an increased risk was once again seen in the highest quintile of ALAN for NHL (HR
2.02, 95% CI 1.03-3.97, Table 3.7). None of the sleep measures were significantly associated with NHL.
3.4. Discussion
Artificial light at night was associated with a 74% increased risk of DLBCL (95% CI 1.13-2.70) and
30% increased risk of NHL overall (95% CI 1.06-1.59). We also observed a non-statistically significant
increased risk of CLL/SLL (HR 1.23, 95% CI 0.83-1.83). There was an association between ALAN and sleep
quality as well as amount of time slept. However, we did not observe an association between any self-
reported measures of sleep and NHL risk. Animal models have shown ALAN induced melatonin
61

suppression can disrupt several circadian regulated cancer pathways, including anti-inflammatory
pathways such as nuclear factor kappa-B (NF-κB) (Blask et al., 2014; Nabavi et al., 2018). NF-κB activates
several inflammatory pathways, including tumor necrosis factor alpha and CXC-chemokine ligand 13,
which have been implicated in NHL etiology (Makgoeng et al., 2018; Sun, 2017). A very small segment of
the cohort (n=5,120) reported sleeping with a light on at night. Of these 5,120 participants, 40 were
diagnosed with NHL, and no association was seen between sleeping with a light on and NHL (HR 0.97,
95% CI 0.71-1.34, Table 3.5). If sleep disruption caused by an indoor source of artificial light is
associated with NHL, we would have expected to see a strong association in this group, but our sample
size was very limited.
There was a modest association between residing in the highest quintile of ALAN and poor sleep
quality (OR 1.16, 95% CI 1.06-1.28, Table 3.6), sleeping less than 8 hours a night (OR 1.22, 95% CI 1.15-
1.30) and reporting having trouble sleeping 1 or more times a week (OR 1.14, 95% CI 1.07-1.22). There
did not seem to be an association between ALAN and taking longer to fall asleep. We are not measuring
light inside the household, which may be a much stronger modifier of sleep onset, especially in more
recent years (Hysing et al., 2015).
Of the few previously reported associations with NHL, we saw increased risk among higher SES
and those with a family history of hematologic malignancies, and decreased risk among current drinkers
(Cerhan, Kricker, et al., 2014; Linet et al., 2014; Morton et al., 2014). We did not see an association with
other reported risk factors, such as pesticide, smoking, or recreational physical activity. Neither alcohol,
SES, nor family history of NHL were confounders in the multivariable model, however physical activity
did modify the HR by more than 10%. While meeting the American Heart Association (AHA)
recommended recreational physical activity levels was not associated with NHL overall (HR 1.05, 95% CI
0.91-1.20), there did appear to be heterogeneity between subtypes with decreasing risk for DLBCL (HR
62

0.92, 95% CI 0.70-1.21) and increasing risk for FL (HR 0.92-1.80). Some research has shown a protective
effect of physical activity and NHL risk (Boyle et al., 2015; Etter et al., 2018; Moore et al., 2016), though
a recent meta-analysis concluded there was no association (Psaltopoulou et al., 2019). Participants
living in higher areas of ALAN were less likely to meet AHA recommended physical activity levels (OR
per
quintile
1.08, 95% CI 1.06-1.09). The risk of NHL in the highest quintile was slightly attenuated (HR 1.31,
95% CI 1.06-1.64) after including physical activity and BMI in the multivariate model. Exercise has been
shown to be a modulator of sleep (Kovacevic, Mavros, Heisz, & Fiatarone Singh, 2018), possibly in
connection with its anti-inflammatory effects (Hojman, Gehl, Christensen, & Pedersen, 2018; Koelwyn,
Wennerberg, Demaria, & Jones, 2015). It is also possible that we are measuring some indication of the
built environment that is having an effect on health.
Strengths of our study include the large sample size and length of follow-up. Our linkage to the
California Cancer registry provides high quality case ascertainment. We also have data on individual
level covariates, such as BMI, alcohol, and exercise. Unfortunately, like many cohort studies, the
interest on sleep disruption and ALAN is relatively recent and that data was not collected at baseline.
The New World Atlas does have its shortcomings when assessing exposure to ALAN. First, it only
estimates brightness at the zenith and does not incorporate light from the whole hemisphere and
especially on the horizon, which drives light exposures in human settlements (Simons et al. submitted).
Second, there is no distinction between the different wavelengths of light. Light in the blue spectrum
(500-450 nm) has been shown to be especially disruptive to sleep (Mortazavi et al., 2018), including
suppression of melatonin (Tähkämö, Partonen, & Pesonen, 2019). A recently published study by MCC-
Spain has used high resolution color photography from the International Space Station to help address
this limitation, though availability of such imagery is limited to larger metropolitan areas(Garcia-Saenz et
al., 2018). We do not have any information on handheld device usage in the cohort; which have been
shown to delay onset of sleep (Mortazavi et al., 2018), however, our study period largely predates the
63

widespread adoption of such devices. While the New World Atlas is a better measure of exposure to
ALAN (Simons et al, submitted for publication), it is based on the 2015 annual VIIRS average of ALAN.
We saw high correlation between the various satellites over time (Table 3.1) and it is likely that in more
developed countries the variation over time is smaller. We believe many of these shortcomings would
cause non-differential misclassification, thereby attenuating any effects; such as those shown in Table
3.5.
As we are continuing to better understand the effects of sleep disruption and built environment,
further studies assessing their risk associated with NHL are warranted.
64

Table 3.1 Correlation of US Defense Meteorological Satellite Operational Linescan System and Visible Infrared Imaging Radiometer Suite
Day/Night Band

DSMP-OLS VIIRS

2006 2011 2012* 2015 2017*
DSMP-OLS 1996 0.94 0.92 0.69 0.68 0.69

2006

0.97 0.72 0.7 0.72

2011

0.72 0.7 0.73
VIIRS 2012*

0.96 0.97

2015

0.98

World Atlas Jittered VIIRS 2015 VIIRS 2015 @ 2.7km DSMP 1996

World Atlas 0.996 0.777 0.823 0.908

World Atlas Jittered

0.769 0.816 0.915

VIIRS 2015

0.887 0.739

VIIRS 2015 @ 2.7km

0.799

* Average of available monthly data without stray light correction

65

Table 3.2 Number of unique geocoded addresses in the California Teachers Study
Unique Geocode Frequency Number of Participants
1 105968 105968
2 1518 3036
3 339 1017
4 170 680
5 90 450
6 67 402
7 42 294
8 28 224
9 19 171
10 24 240
11 7 77
12 6 72
13 6 78
14 6 84
15 5 75
16 2 32
17 3 51
18 1 18
19 1 19
20 1 20
23 2 46
24 3 72
25 1 25
26 1 26
27 1 27
28 1 28
30 1 30
32 1 32
34 2 68
36 1 36
43 1 43

66

Table 3.3 Demographic characteristics of the California Teachers Study Cohort in light at night analysis

Cohort NHL

n=105,486 n=933
Race

White 91,987 87.2% 841 90.1%
Black 2,474 2.3% 17 1.8%
Hispanic 4,603 4.4% 22 2.4%
Asian/Pacific islander 4,470 4.2% 36 3.9%
Other 1,952 1.9% 17 1.8%

Light at Night

Median (IQR) 2.40 (1.22-3.95) 2.50 (1.27-4.29)
Quintile 1 21,296 20.2% 174 18.6%
Quintile 2 21,294 20.2% 187 20.0%
Quintile 3 21,183 20.1% 183 19.6%
Quintile 4 20,959 19.9% 175 18.8%
Quintile 5 20,754 19.7% 214 22.9%

UV

Quintile 1 18,014 20.0% 154 19.5%
Quintile 2 18,270 20.3% 158 20.0%
Quintile 3 17,616 19.6% 186 23.5%
Quintile 4 18,224 20.2% 177 22.4%
Quintile 5 17,945 19.9% 116 14.7%

SES

Quartile 1 4,573 4.4% 24 2.6%
Quartile 2 18,206 17.5% 137 14.8%
Quartile 3 34,391 33.0% 289 31.2%
Quartile 4 46,925 45.1% 476 51.4%

Rural Census Block

Rural 14,953 14.4% 118 12.7%
Suburban 22,989 22.1% 176 19.0%
Urban 66,197 63.6% 632 68.3%

BMI

<20 11,264 11.1% 79 8.9%
20-24.9 51,289 50.6% 428 48.3%
25-29.9 24,911 24.6% 246 27.7%
30+ 13,859 13.7% 134 15.1%

Smoking

67

Never 70,275 67.0% 575 62.0%
Former 29,393 28.0% 311 33.5%
Current 5,156 4.9% 41 4.4%

Alcohol

Never 33,441 33.4% 302 33.5%
Former 58,316 58.3% 534 59.3%
Current 8,285 8.3% 65 7.2%

Any Daily NSAID

No 73,007 70.2% 642 69.9%
Yes 30,968 29.8% 276 30.1%

Exercise

Did not meet AHA guidelines 35,845 34.2% 299 32.5%
Met AHA guidelines 68,900 65.8% 622 67.5%

Family History of NHL

No 96,745 95.2% 833 92.0%
Yes 4,872 4.8% 72 8.0%

Any exposure to pesticides

No 66,285 62.8% 634 68.0%
Yes 39,201 37.2% 299 32.0%

Histology

DLBCL

234

FL

179

CLL/SLL

229

Marginal Zone

89

Other

202

68

Table 3.4 Associations between light at night and select demographic factors with non-Hodgkin lymphoma

Overall NHL (n=933) DLBCL (n=234) FL (n=179) CLL/SLL (n=229)
Age adjusted

Light at Night HR 95% CI HR 95% CI HR 95% CI HR 95% CI
Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.13 (0.92-1.40) 1.49 (0.95-2.33) 1.34 (0.86-2.09) 0.94 (0.61-1.43)
Quintile 3 1.08 (0.88-1.34) 1.48 (0.94-2.32) 1.13 (0.71-1.80) 0.85 (0.55-1.31)
Quintile 4 1.02 (0.83-1.27) 1.39 (0.88-2.19) 0.84 (0.51-1.38) 0.98 (0.65-1.49)
Quintile 5 1.30 (1.06-1.59) 1.74 (1.13-2.70) 0.86 (0.52-1.41) 1.23 (0.83-1.83)

UV

Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.09 (0.87-1.36) 1.06 (0.66-1.72) 1.21 (0.75-1.97) 0.90 (0.58-1.40)
Quintile 3 1.25 (1.01-1.56) 1.43 (0.92-2.24) 0.92 (0.55-1.56) 1.05 (0.69-1.59)
Quintile 4 1.24 (0.99-1.54) 1.66 (1.07-2.57) 0.95 (0.56-1.60) 0.94 (0.61-1.44)
Quintile 5 0.88 (0.68-1.12) 1.08 (0.66-1.76) 1.02 (0.61-1.72) 0.56 (0.33-0.94)

Census Block SES

Quartile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quartile 2 1.50 (0.95-2.35) 2.18 (0.66-7.22) 1.60 (0.62-4.10) 3.88 (0.93-16.2)
Quartile 3 1.67 (1.08-2.58) 3.24 (1.02-10.3) 1.19 (0.47-2.99) 4.76 (1.17-19.4)
Quartile 4 1.81 (1.18-2.78) 3.53 (1.12-11.1) 1.48 (0.60-3.64) 4.81 (1.19-19.5)

BMI

<20 0.97 (0.76-1.24) 1.22 (0.77-1.92) 0.69 (0.36-1.34) 1.29 (0.84-2.00)
20-24.9 1.00 - 1.00 - 1.00 - 1.00 -
25-29.9 1.05 (0.90-1.24) 1.04 (0.75-1.44) 1.25 (0.87-1.79) 0.86 (0.62-1.20)
30+ 1.12 (0.92-1.36) 1.32 (0.90-1.92) 1.41 (0.92-2.16) 0.66 (0.41-1.07)

Smoking

Never 1.00 - 1.00 - 1.00 - 1.00 -
69

Former 1.04 (0.90-1.19) 1.02 (0.77-1.36) 0.80 (0.57-1.12) 1.03 (0.78-1.36)
Current 0.91 (0.66-1.26) 1.09 (0.60-1.97) 0.89 (0.44-1.83) 0.45 (0.18-1.10)

Alcohol

Never 1.00 - 1.00 - 1.00 - 1.00 -
Former 0.97 (0.84-1.12) 0.84 (0.64-1.12) 0.93 (0.67-1.29) 1.32 (0.97-1.79)
Current 0.73 (0.55-0.95) 0.72 (0.43-1.23) 0.86 (0.48-1.54) 0.91 (0.52-1.58)

Any Daily NSAID

no 1.00 - 1.00 - 1.00 - 1.00 -
yes 0.92 (0.80-1.07) 0.77 (0.57-1.04) 1.01 (0.73-1.40) 0.88 (0.66-1.18)

Exercise

did not meet AHA guidelines 1.00 - 1.00 - 1.00 - 1.00 -
met AHA guidelines 1.05 (0.91-1.20) 0.92 (0.70-1.21) 1.29 (0.92-1.80) 0.97 (0.73-1.28)

Family History of NHL

no 1.00 - 1.00 - 1.00 - 1.00 -
yes 1.49 (1.17-1.90) 1.40 (0.85-2.29) 1.80 (1.07-3.01) 1.41 (0.86-2.31)

Any exposure to pesticides

no 1.00 - 1.00 - 1.00 - 1.00 -
yes 1.01 (0.87-1.16) 1.03 (0.77-1.37) 1.07 (0.77-1.48) 0.77 (0.57-1.05)

Age and race adjusted Overall NHL (n=933) DLBCL (n=234) FL (n=179) CLL/SLL (n=229)
Light at Night HR 95% CI HR 95% CI HR 95% CI HR 95% CI
Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.14 (0.93-1.41) 1.50 (0.96-2.35) 1.36 (0.87-2.12) 0.94 (0.62-1.44)
Quintile 3 1.10 (0.89-1.36) 1.50 (0.96-2.35) 1.17 (0.74-1.86) 0.86 (0.56-1.33)
Quintile 4 1.04 (0.84-1.29) 1.41 (0.90-2.23) 0.87 (0.53-1.44) 1.00 (0.66-1.51)
Quintile 5 1.35 (1.10-1.66) 1.81 (1.17-2.82) 0.93 (0.56-1.54) 1.27 (0.85-1.89)
70

Age, race, and physical activity
adjusted Overall NHL (n=933) DLBCL (n=234) FL (n=179) CLL/SLL (n=229)
Light at Night HR 95% CI HR 95% CI HR 95% CI HR 95% CI
Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.24 (0.99-1.57) 1.70 (1.03-2.81) 1.37 (0.85-2.22) 0.95 (0.59-1.52)
Quintile 3 1.15 (0.91-1.45) 1.78 (1.08-2.93) 1.14 (0.69-1.88) 0.81 (0.49-1.32)
Quintile 4 1.10 (0.87-1.40) 1.66 (1.00-2.76) 0.72 (0.41-1.27) 1.05 (0.67-1.66)
Quintile 5 1.44 (1.15-1.81) 2.03 (1.24-3.33) 0.87 (0.50-1.51) 1.34 (0.86-2.08)

Multivariate age, race, BMI, and
physical activity adjusted Overall NHL (n=933) DLBCL (n=234) FL (n=179) CLL/SLL (n=229)
Light at Night HR 95% CI HR 95% CI HR 95% CI HR 95% CI
Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.14 (0.92-1.42) 1.47 (0.93-2.33) 1.42 (0.88-2.28) 0.99 (0.63-1.54)
Quintile 3 1.12 (0.90-1.39) 1.46 (0.93-2.31) 1.31 (0.81-2.13) 0.85 (0.53-1.35)
Quintile 4 1.02 (0.81-1.27) 1.33 (0.83-2.11) 0.91 (0.53-1.55) 0.99 (0.64-1.55)
Quintile 5 1.40 (1.13-1.73) 1.76 (1.13-2.75) 1.05 (0.62-1.77) 1.36 (0.89-2.07)

Multivariate age, race, SES, BMI,
and physical activity adjusted Overall NHL (n=933) DLBCL (n=234) FL (n=179) CLL/SLL (n=229)
Light at Night HR 95% CI HR 95% CI HR 95% CI HR 95% CI
Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.08 (0.86-1.35) 1.31 (0.93-2.07) 1.38 (0.85-2.25) 0.92 (0.59-1.45)
Quintile 3 1.05 (0.83-1.31) 1.29 (0.81-2.04) 1.28 (0.78-2.12) 0.78 (0.49-1.26)
Quintile 4 0.96 (0.76-1.20) 1.23 (0.77-1.95) 0.89 (0.51-1.53) 0.92 (0.59-1.45)
Quintile 5 1.32 (1.06-1.64) 1.58 (1.01-2.46) 1.03 (0.60-1.76) 1.27 (0.83-1.95)
71

Table 3.5 Additional sensitivity analyses for association between light at night and non-Hodgkin lymphoma

Overall NHL (n=933)
Age adjusted World Atlas VIIRS VIIRS @ 2.7km DSMP-OLS
Light at Night HR 95% CI HR 95% CI HR 95% CI HR 95% CI
Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.13 (0.92-1.40) 1.12 (0.91-1.37) 1.07 (0.87-1.32) 1.23 (1.00-1.52)
Quintile 3 1.08 (0.88-1.34) 1.19 (0.97-1.46) 1.11 (0.90-1.37) 1.05 (0.85-1.31)
Quintile 4 1.02 (0.83-1.27) 1.00 (0.81-1.24) 1.16 (0.95-1.43) 1.10 (0.89-1.36)
Quintile 5 1.30 (1.06-1.59) 1.08 (0.87-1.33) 1.11 (0.90-1.37) 1.21 (0.98-1.49)

Age adjusted Overall NHL (n=933) DLBCL (n=234) FL (n=179) CLL/SLL (n=229)
Light at Night HR 95% CI HR 95% CI HR 95% CI HR 95% CI
Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.13 (0.92-1.40) 1.49 (0.95-2.33) 1.34 (0.86-2.09) 0.94 (0.61-1.43)
Quintile 3 1.08 (0.88-1.34) 1.48 (0.94-2.32) 1.13 (0.71-1.80) 0.85 (0.55-1.31)
Quintile 4 1.02 (0.83-1.27) 1.39 (0.88-2.19) 0.84 (0.51-1.38) 0.98 (0.65-1.49)
Quintile 5 1.30 (1.06-1.59) 1.74 (1.13-2.70) 0.86 (0.52-1.41) 1.23 (0.83-1.83)

Age and exercise adjusted

Light at Night

Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 1.28 (1.01-1.62) 1.64 (0.99-2.72) 1.52 (0.93-2.49) 1.03 (0.64-1.66)
Quintile 3 1.15 (0.90-1.46) 1.61 (0.97-2.68) 1.26 (0.75-2.11) 0.88 (0.53-1.45)
Quintile 4 1.09 (0.86-1.40) 1.61 (0.97-2.68) 0.79 (0.44-1.41) 1.09 (0.68-1.75)
Quintile 5 1.45 (1.15-1.82) 1.88 (1.14-3.09) 0.92 (0.52-1.61) 1.44 (0.92-2.25)

Age, race, and exercise adjusted

Light at Night

Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
72

Quintile 2 1.17 (0.95-1.44) 1.50 (0.96-2.35) 1.44 (0.92-2.28) 0.96 (0.63-1.47)
Quintile 3 1.11 (0.90-1.38) 1.50 (0.95-2.34) 1.24 (0.77-1.99) 0.83 (0.53-1.29)
Quintile 4 1.05 (0.85-1.31) 1.41 (0.90-2.22) 0.93 (0.56-1.55) 1.02 (0.67-1.55)
Quintile 5 1.38 (1.13-1.70) 1.80 (1.16-2.80) 1.00 (0.60-1.67) 1.30 (0.87-1.95)

Overall NHL (n=40) DLBCL (n=8) FL (n=5) CLL/SLL (n=9)
Sleeping with light in bedroom 0.97 (0.71-1.34) 0.80 (0.40-1.62) 0.64 (0.26-1.55) 0.93 (0.48-1.81)

Light at Night age and race adjusted

Bottom 5% 0.92 (0.67-1.26)

Middle 50% 1.00 -

Top 5% 1.39 (1.04-1.85)

73

Table 3.6 Self-reported sleep quality and select baseline demographics among CTS participants
reporting same sleeping habits over the past year

Cohort (n=40,613) NHL (n=91)
Age at questionnaire 5 (mean, sd) 67.2 10.9 72.8 8.9

Race

White 36042 88.7% 80 87.9%
Black 667 1.6% 0 0.0%
Hispanic 1585 3.9% 2 2.2%
Asian/Pacific islander 1714 4.2% 7 7.7%
Other 605 1.5% 0 0.0%

Light at Night

Quintile 1 8,384 20.6% 14 15.4%
Quintile 2 8,313 20.5% 19 20.9%
Quintile 3 8,244 20.3% 13 14.3%
Quintile 4 7,983 19.7% 21 23.1%
Quintile 5 7,689 18.9% 24 26.4%

Chronotype

Morning 17473 43.0% 38 41.8%
More morning than evening 6322 15.6% 18 19.8%
Neither 5533 13.6% 16 17.6%
More evening than morning 5349 13.2% 9 9.9%
Evening 5936 14.6% 10 11.0%

Sleep Quality

Very good 12790 31.5% 30 33.0%
Fairly good 22129 54.5% 50 54.9%
Fairly poor 5095 12.5% 9 9.9%
Very poor 442 1.1% 1 1.1%

Trouble Falling Asleep During Past Month

No 9216 22.7% 26 28.6%
<1 times/week 13639 33.6% 26 28.6%
1-2 times/week 10082 24.8% 27 29.7%
3+ times/week 7600 18.7% 12 13.2%

Time takes to fall asleep

<15 min 19579 48.2% 42 46.2%
16-30 min 14575 35.9% 34 37.4%
31-60 min 4901 12.1% 13 14.3%
74

>60 min 1421 3.5% 2 2.2%

Time slept

<5 hours 927 2.3% 1 1.1%
5-6 hours 9420 23.2% 18 19.8%
7 hours 17256 42.5% 43 47.3%
8 hours 10841 26.7% 21 23.1%
>9 hours 1917 4.7% 6 6.6%

Use of medicine during past month

No 28648 70.5% 68 74.7%
<1 times/week 4384 10.8% 10 11.0%
1-2 times/week 2360 5.8% 3 3.3%
3+ times/week 5221 12.9% 10 11.0%

SES

Quartile 1 1480 3.6% 1 1.1%
Quartile 2 6364 15.7% 11 12.1%
Quartile 3 13234 32.6% 34 37.4%
Quartile 4 19051 46.9% 45 49.5%

Alcohol

Never 11968 29.5% 28 30.8%
Former 23886 58.8% 59 64.8%
Current 3240 8.0% 2 2.2%

Family History of NHL

no 37592 92.6% 82 90.1%
yes 1855 4.6% 5 5.5%
75

Table 3.7 Association between self-reported sleep quality and artificial light at night among CTS
participants reporting same sleeping habits over the past year (n=40,613)

Univariate Multivariate*
Very Good/World Atlas Quintile 1

1.00 - 1.00 -
Fairly Good

World Atlas Quintile 2

1.00 (0.93-1.06) 0.99 (0.93-1.05)
World Atlas Quintile 3

0.97 (0.91-1.04) 0.97 (0.91-1.03)
World Atlas Quintile 4

1.02 (0.96-1.09) 1.02 (0.96-1.09)
World Atlas Quintile 5

1.02 (0.96-1.09) 1.02 (0.95-1.09)
Fairly Poor/Very Poor

World Atlas Quintile 2

1.12 (1.02-1.23) 1.10 (1.00-1.20)
World Atlas Quintile 3

1.04 (0.95-1.14) 1.02 (0.93-1.12)
World Atlas Quintile 4

1.15 (1.05-1.26) 1.12 (1.02-1.24)
World Atlas Quintile 5

1.20 (1.09-1.32) 1.16 (1.06-1.28)

Sleeping <8 hours/night

World Atlas Quintile 1

1.00 - 1.00 -
World Atlas Quintile 2

1.07 (1.00-1.13) 1.05 (0.99-1.12)
World Atlas Quintile 3

1.16 (1.09-1.24) 1.13 (1.06-1.20)
World Atlas Quintile 4

1.20 (1.12-1.27) 1.16 (1.09-1.24)
World Atlas Quintile 5

1.31 (1.23-1.40) 1.22 (1.15-1.30)

Taking >15 min to fall asleep**

World Atlas Quintile 1

1.00 - 1.00 -
World Atlas Quintile 2

0.95 (0.90-1.01) 0.94 (0.89-1.00)
World Atlas Quintile 3

0.96 (0.91-1.02) 0.96 (0.90-1.02)
World Atlas Quintile 4

0.96 (0.90-1.01) 0.94 (0.88-1.00)
World Atlas Quintile 5

1.00 (0.94-1.05) 0.98 (0.92-1.04)

Trouble sleeping >1 times/week

World Atlas Quintile 1

1.00 - 1.00 -
World Atlas Quintile 2

1.07 (1.00-1.14) 1.07 (1.00-1.14)
World Atlas Quintile 3

1.04 (0.97-1.10) 1.04 (0.98-1.11)
World Atlas Quintile 4

1.12 (1.05-1.20) 1.12 (1.05-1.20)
World Atlas Quintile 5

1.14 (1.07-1.22) 1.14 (1.07-1.22)

*Adjusted for age, race, and chronotype
**Adjusted for age, race, chronotype, and use of sleep medication

76

Table 3.8 Association with light at night and non-Hodgkin lymphoma with follow-up beginning at time
of response to questionnaire 5 (2012-2013)
Light at Night HR 95% CI
Quintile 1 1.00 -
Quintile 2 1.38 (0.68-2.83)
Quintile 3 0.87 (0.39-1.94)
Quintile 4 1.72 (0.86-3.43)
Quintile 5 2.02 (1.03-3.97)

Sleep Quality

Very good 1.00 -
Fairly good 1.07 (0.67-1.70)
Very/fairly poor 0.98 (0.48-2.03)

Time Slept

8+ hours 1.00 -
7 hours 0.98 (0.54-1.77)
<6 hours 1.22 (0.74-2.00)

Time takes to fall asleep

<15 minutes 1.00 -
16-30 minutes 1.05 (0.66-1.67)
> 30 minutes 1.18 (0.65-2.14)

Alcohol

Never 1.00 -
Former 0.92 (0.59-1.46)
Current 0.22 (0.05-0.93)

Census Block SES

Quartile 1 1.00 -
Quartile 2 2.28 (0.29-17.8)
Quartile 3 3.53 (0.48-25.8)
Quartile 4 2.72 (0.37-19.7)

Family History of NHL

no 1.00 -
yes 1.06 (0.43-2.63)

77

Figure 3.1 Exclusion/inclusion criteria for California Teachers Study participants in analysis of artificial
light at night and cancer

78

Figure 3.2 Distribution of the New World Atlas of Artificial Sky Brightness in the California Teachers
Study cohort with quintile cutoffs

Brightness (mcd/m
2
)
79

Figure 3.3 Number of California Teachers Study Participants by County

80

Figure 3.4 Median age at study entry of California Teachers Study participants by county

81

Figure 3.5 Number of non-Hodgkin lymphoma cases in the California Teachers Study cohort by county

82

Figure 3.6 Distribution of New World Atlas of Artificial Night Sky Brightness in the California Teachers
Study cohort

Quartile 1
Quartile 3
Median
New World Atlas Night at Light
intensity (mcd/m
2
)
83

Figure 3.7 Self-reported sleep questions from the California Teachers Study wave V questionnaire

84

Chapter 4. Air Pollution and Risk of Non-Hodgkin Lymphoma in the California Teachers Study Cohort
4.1 Introduction
Outdoor environmental air pollution (AP) is made of a heterogeneous mixture of solid and semi-
volatile materials of various sizes. Over the last decade, over a dozen studies in the US have
demonstrated associations between daily exposure to fine particles (particulate matter less than 2.5
microns in aerodynamic diameter or PM
2.5
) and premature mortality from cardiovascular disease
(Brunekreef & Holgate, 2002). It is estimated that approximately 3.7 million deaths in 2012 were
attributed to poor air quality(Kelly & Fussell, 2015).
Several conflicting studies have been published on air pollution and NHL risk. Studies conducted
among US veterans and among SEER-Medicare beneficiaries found an association between
hospitalization for acute inflammatory conditions, such as chronic obstructive pulmonary disease and
pneumonia, with increased NHL risk (Anderson et al., 2014; Koshiol et al., 2011). Acute exposure to
poor air quality has also been linked to increased hospital admissions for respiratory conditions such as
pneumonia (Zhang, Hong, & Liu, 2017). Mice exposed to PM
2.5
had higher level of circulating
inflammatory cytokines, including tumor necrosis factor (TNF), compared to mice exposed to filtered air
(Xu et al., 2013). Increases in PM
2.5
has been associated with higher circulating markers of inflammation
such as C-reactive protein (CRP), interlukein-6 (IL-6), interlukein-10 (IL-10), and TNF (Gruzieva et al.,
2017; Li et al., 2017). In the Multi-Ethnic Study of Atherosclerosis, a six percent increase in IL-6 was seen
per 5 µg/m
3
increase in PM
2.5
(Hajat et al., 2015). These pathways of increased inflammation have been
associated with increased risk of developing NHL (Conroy et al., 2013; Edlefsen et al., 2014; Elena
Vendrame et al., 2014). More recent interest has turned to ultrafine particles, those under 0.1 microns
(PM
0.1
). It is believed that these smaller particles more readily enter the blood stream. Studies have
85

shown inflammatory effects of exposure to these smaller particles (Donaldson et al., 2002; Donaldson,
Stone, Clouter, Renwick, & MacNee, 2001; Kreyling, Semmler, & Moller, 2004; Oberdorster, 2000).
The relationship between long-term exposures to low levels of immune modulating pollutants
with NHL, however, has not been well-studied, but links to adverse health outcomes are emerging
(Ostro et al., 2015). A systematic review of occupational exposure in commercial drivers found elevated
risk of Hodgkin’s lymphoma (SMR 2.17, 95% CI 1.19-3.87) in bus drivers compared to white-collar
workers. Living in close proximity to AP from traffic and industrial factories has been associated with
other hematologic malignancies, such as leukemia in children (Magnani et al., 2016; Parodi et al., 2015).
Mixed results however have been seen for NHL risk in those living in close proximity to industrial sources
of AP, with either no risk or increased risk (De Roos et al., 2010; Parodi et al., 2014; Ramis et al., 2009).
While long-term exposure to PM
2.5
has been associated with lung cancer (Loomis et al., 2013; Pun,
Kazemiparkouhi, Manjourides, & Suh, 2017; M. C. Turner et al., 2017), results have been mixed for other
cancers, including NHL (Hung, Chan, Wu, Chiu, & Yang, 2012; Loomis et al., 2013; M. C. Turner et al.,
2017; Yeh et al., 2017). However, many of these studies often involve the use of a single central city
monitor or distance from a point source to measure pollution concentrations for an entire metropolitan
area and lack personal level covariates, resulting in a significant degree of misclassification of exposure
and biased effect estimates. In this chapter, we utilized a source-oriented transport model of air
pollution that estimates exposure to PM
0.1
as well as PM
2.5
(J. Hu, Zhang, et al., 2014a; J. Hu, Zhang, et
al., 2014b) and NHL risk in the prospective California Teachers Study (CTS) cohort.
4.2 Methods
4.2.1 Study Population
We will once again utilize the CTS to evaluate the association between air pollution and NHL
risk. Participants were excluded if they: lived outside of California at baseline or could not be geocoded
86

(n=14,427), had prevalent cancer (n=8,078), we were unable to assign air pollution (n=7,924), or ended
follow-up prior to January 1, 2000 (due to the availability of air pollution measures, n=2,189) for a final
analytic cohort of 100,861 (Figure 4.1).
4.2.2 Exposure Assessment
Historically, studies looking at disease associations and air pollution utilized data from
monitoring stations. However, the scarcity of these stations meant that a single observation would
be used for an entire city or metropolitan area. Use of spatial modeling techniques have been used
to better assign exposure, including in the CTS (Lipsett et al., 2011; Ostro et al., 2010), though these
methods are limited the further one goes from the monitoring station. Of the 133,000 CTS
participants, approximately 60% resided within a 10km radius of a monitoring station. In this
analysis, we utilized various source-oriented chemical transport models (CTM), which were able to
estimate exposure for over 90% of the cohort.
4.2.2.1 Harvard Model
The Harvard air pollution model augments a CTM by including additional data from satellites (Di
et al., 2016). The CTM the model is built upon is the GEOS-Chem model. GEOS-Chem is a global CTM
based on emissions inventories and meteorological observations from the Goddard Earth Observing
System (GEOS). To improve upon the base model, weather patterns were calculated using the National
Centers for Environmental Prediction North American Regional Reanalysis data. Aerosol optical depth
(AOD) was measured daily through the Moderate Resolution Imaging Spectroradiometer (MODIS)
satellite. AOD measures the amount of sunlight that is blocked or scattered by particles in the
atmosphere. Through algorithms such as the Multi-Angle Implementation of Atmospheric Correction
(MAIAC), that was utilized in this model, it is able to provide an estimate of PM
2.5
at a 1km resolution
(Lyapustin, Martonchik, Wang, Laszlo, & Korkin, 2011). MODIS surface reflectance data was included as
87

well to adjust for surface reflectance, which can impact the reflectance of light, and the measured AOD.
Also included was the absorbing aerosol index, to account for the particulates that can scatter and
absorb sunlight. Additional land use variables included elevation, road density, National Emissions
Inventory, population density, urbanization, and normalized difference vegetation index (NDVI). A
neural network was trained to predict PM2.5 with these data at monitoring stations. Daily PM
2.5
mass
was modeled on a national level. We subset the daily values for the state of California and created
monthly averages for use in our analyses.
4.2.2.2 Kleeman Model
The University of California, Davis/California Institute of Technology source oriented CTM,
hereafter referred to as the Kleeman model, was developed specifically for the state of California.
This model utilizes the extensive emissions inventory in California provided by the California Air
Resources Board (CARB), which tracks over 900 sources, to model pollution concentrations (specifically,
mass and number concentrations of PM constituents in particle diameters ranging from 0.01 to 10 µm)
through calculations that describe transport, diffusion, deposition, coagulation, gas- and particle-phase
chemistry, and gas-to-particle conversion (J. Hu, Zhang, et al., 2014b; J. Hu, Zhang, Ying, et al., 2014). In
addition, they include data from the National Center for Atmospheric Research Fire Inventory to track
particulate matter from wildfires. Meteorological fields were generated with the Weather Research and
Forecasting (WRF) model using data assimilation for measured meteorological parameters where
available. The data is then provided on a monthly scale for PM
0.1
and PM
2.5
across the state in a 4km
grid. Uniquely, data on various constituents are available through this model; however, the model is
currently undergoing revisions and was therefore included in exploratory analyses focused particularly
on organic carbons, which have been shown to be associated with NHL (DellaValle et al., 2016;
Mastrangelo, Fadda, & Marzia, 1996).
88

4.2.3 Outcome Assessment
Our cases were obtained through linkage to the California Cancer Registry. Non-Hodgkin
lymphoma (NHL) was defined as having a diagnosis with the International Classification of Diseases for
Oncology, third edition codes of: diffuse large B-cell lymphoma (DLBCL: 9678, 9679, 9680, 9684),
follicular lymphoma (FL: 9690, 9691, 9695, 9698), mantle cell lymphoma (MCL: 9673), marginal zone
lymphoma (MZL: 9699), chronic lymphocytic leukemia/Small lymphocytic lymphoma (CLL/SLL: 9670,
9823), Burkitt lymphoma (BL: ICD-O-3 codes 9687, 9826), and other B cell lymphomas (other, including
not otherwise specified (NOS): 9590, 9591, 9596, 9671, 9675, 9727, 9728, 9833, 9835, 9836, 9761). We
identified a total of 762 incident cases of NHL in the cohort from January 1, 2000 through December 31,
2016.
4.2.4 Statistical Methods
We evaluated the association between air pollution and NHL overall and among the three most
common subtypes (DLBCL, FL, CLL/SLL). We calculated hazard ratios (HR) and 95% confidence intervals
(95% CI) using a Cox proportional hazards model. In the Cox regression models, the time scale (in
months) was defined by age at entry into the cohort and the first of the following ages: at event (NHL
diagnosis), at censoring (e.g., when a participant moved out of California for more than four months), at
death, or at end of follow-up (December 31, 2016). Air pollution was measured using a time-dependent
average starting at January 1, 2000 until month of follow-up. We assessed confounding among several
covariates including; age, race (non-Hispanic White, Black, Hispanic, Asian/Pacific Islander, Other),
alcohol and smoking status (current, former, never), body mass index (<18.5, 18.5-25, 25-30, or ≥30
kg/m
2
), United States Census block level socioeconomic status (statewide quartiles), United States
Census block urban/rural status (Ratcliffe et al., 2016), use of non-steroidal anti-inflammatory drugs,
exercise (meeting/not meeting American Heart Association recommended physical activity levels
89

(Troiano, 2018)), exposure to pesticides, hormone use, and family history of hematologic malignancy.
All analyses were conducted in SAS 9.4 (SAS Institute Inc., Cary, NC).
4.3 Results
4.3.1 Harvard Air Pollution Model
The cohort was overall similar to that in the light at night analysis in chapter 3; with the majority
being non-Hispanic White and of high SES (Table 4.1). Of the 100,861 eligible participants, we were able
to assign air pollution measures to 100,103 across the follow-up period. The CTS cohort has low
prevalence of smoking, with just 5% at reporting being current smokers at baseline.
The mean of PM
2.5
over the study period was 12.3 μg/m
3
with a standard deviation of 3.4 μg/m
3

and a range of 2.75 μg/m
3
– 26.48 μg/m
3
. The highest levels of PM
2.5
were in the central valley (Figure
4.3). Overall, the average level of PM
2.5
has decreased over time in the cohort from a high in 2001 of
15.58 μg/m
3
to the current levels of 9.99 μg/m
3
in 2016 (Figure 4.2). There was not a statistically
significant increase in risk of NHL per 10 μg/m
3
of PM
2.5
(HR 1.06, 95% CI 0.89-1.27). The association was
slightly stronger, though not statistically significant, with DLBCL (HR 1.07, 95% CI 0.75-1.53) and CLL (HR
1.08, 95% CI 0.75-1.57). There appeared to be a protective effect for FL (HR 0.87, 95% CI 0.68-1.33).
The results were similar after adjusting for age, race, SES, and smoking status with the overall risk of NHL
increasing by 11% (95% CI 0.93-1.33) per 10 μg/m
3
increase in PM
2.5
. The risk of DLBCL (HR 1.14, 95% CI
0.80-1.64) and CLL (HR 1.12, 95% CI 0.77-1.63) slightly increased, while the protective effect seen in FL
risk was attenuated (HR 0.93, 95% CI 0.62-1.41).
48.1% of the cohort had an average exposure to PM
2.5
above the nationally recommended level
of 12 μg/m
3
, during their follow-up time. Compared to participants who resided in areas with less than
12 μg/m
3
of PM
2.5
, there was a 19% increased risk of NHL (95% CI 1.03-1.38). The association was again
stronger in DLBCL (HR 1.27, 95% CI 0.94-1.71) and CLL (0.90-1.65). The association with FL was
90

attenuated (HR 0.95, 95% CI 0.68-1.33). Adjusting for age, race, SES, and smoking did not greatly alter
risk.
4.3.2 Kleeman Air Pollution Model
There was only moderate correlation (0.53, Table 4.5) between the Harvard 1km model and
Kleeman 4km model; the mean averaged PM
2.5
for the Kleeman model was 7.73 μg/m
3
with a standard
deviation of 1.89 μg/m
3
(Table 4.3); much lower than the Harvard model. The range was similar with a
minimum of 2.35 μg/m
3
and maximum of 26 μg/m
3
. The mean averaged PM
0.1
across the cohort was
0.64 μg/m
3
with a standard deviation of 0.25. The average was slightly higher in the cases at 0.74
μg/m
3
. The average ranged from no exposure to a maximum of 5.23 μg/m
3
. The highest level of
exposure was in Kern county, with an average of 1.22 μg/m
3
. There was no association with PM
2.5
and
NHL risk (per 10 μg/m
3
; HR 0.97, 95% CI 0.69-1.35), or between PM
0.1
and NHL risk (HR 0.94, 95% CI
0.72-1.22, Table 4.4).
The average exposure to PM
2.5
metals was 0.27 μg/m
3
with a standard deviation of 0.09. There
did not appear to be an association with PM
2.5
metals and NHL risk (adjusted HR 0.97, 95% CI 0.90-1.05).
The average exposure to PM
2.5
organic carbons was 1.34 μg/m
3
with a standard deviation of 0.59,
ranging from 0.17-9.90 μg/m
3
. After adjusting for age, race, and smoking status a 1 μg/m
3
increase of
PM
2.5
organic carbons was associated with a 3% increased risk (95% CI 0.92-1.13) of NHL (Table 4.3).
There appeared to be a statistically significant protective effect for FL (HR 0.75, 95% CI 0.58-0.97) that
was attenuated after adjusting for confounders (HR 0.76, 95% CI 0.58-1.00).
4.3.3 Combined Light at Night and Air Pollution Model
There was a moderate correlation between the Harvard air pollution model and the World Atlas
(0.60) (Table 4.5). In a multivariate model of air pollution and light at night, the effects of both PM
2.5
(HR
91

1.07, 95% CI 0.85-1.36, per 10 μg/m
3
PM
2.5
) and the World Atlas were attenuated (HR 1.13, 95% CI 0.84-
1.51, quartile 1 vs quartile 5).
4.4 Discussion
There was a 19% increased risk of NHL (95% CI 1.03-1.38) for participants who were exposed to
PM
2.5
levels above the EPA air quality standards (12 μg/m
3
). The associations were stronger for DLBCL
(HR 1.32, 95% CI 0.98-1.77) and CLL (HR 1.22, 95% CI 0.90-1.66), though marginally not statistically
significant. We did not see a statistically significant increased risk of NHL per 10 μg/m
3
increase in PM
2.5

(adjusted HR 1.09, 95% CI 0.91-1.30). Air pollution in California has been steadily decreasing due to
stricter emission laws, reducing the amount of exposure to PM over time (Garcia et al., 2019; Lurmann,
Avol, & Gilliland, 2015). The components of tailpipe emissions, such as benzene and polycyclic aromatic
hydrocarbons (PAH) that result from incomplete combustion of organic matter, are those that have
been associated with inflammation (Burchiel & Luster, 2001) and several cancers, including NHL (Darbre,
2018; Mastrangelo et al., 1996; Singh et al., 2018; Smith, Jones, & Smith, 2007). While overall PM levels
have been decreasing with improved emissions laws, the number of vehicles has steadily increased,
causing brake dust and other particulates on the road to become a larger source of PM (Khan & Strand,
2018; McDonald, Goldstein, & Harley, 2015).
When analyzing specific species using the Kleeman air pollution model, there was a non-
significant, 3% decreased risk of NHL (95% CI 0.90-1.05) per 100 ng/m
3
increase in PM
2.5
metals and a 3%
(95% CI 0.93-1.14) increased risk of NHL for each 1 μg/m
3
increase in PM
2.5
organic carbons. The
strongest association was a 13% increased risk of CLL per 1 μg/m
3
increase in PM
2.5
organic carbons (95%
CI 0.93-1.38). The correlation between the Harvard model and the World Atlas (0.60) was stronger than
between the Harvard model and the Kleeman model (0.53). The underestimation of exposure possibly
led to the attenuation of risk; which may also affect our analyses of the species. The version of the
92

Kleeman model presented here is the only one that has been estimated for the study period (2000-
2016), and is known to underestimate exposure. An updated version of model will be available for
analysis in the future (personal communication, M. Kleeman).
Our analysis focused on outdoor air pollution; we do not account for indoor air pollution. The
correlation between outdoor and indoor air pollution can vary based on a variety of factors, such as air
conditioning design and use, building design, and distance from sources of turbulent air such as
roadways (Bell, Ebisu, Peng, & Dominici, 2009; Tong, Chen, Malkawi, Adamkiewicz, & Spengler, 2016).
One study has shown increased risk of NHL from PAH associated with carpet dust, but only for T-cell NHL
(DellaValle et al., 2016). It has been suggested that total exposure, and not just average exposure may
play a role in NHL etiology (Ramis et al., 2009); however both of these exposures are difficult to measure
retrospectively. Our population was also limited to the state of California, with a large portion residing
in Los Angeles County (Figure 3.3), which has been known for historically high levels of air pollution.
Advantages of our study include the long follow-up and large and geographically diverse
population of the CTS. We have accrued a relatively large number of cases in comparison to other
studies of air pollution and NHL (De Roos et al., 2010; Parodi et al., 2014; Ramis et al., 2009). We also
improved upon use of single monitoring stations to assign exposure, and were able to explore both
PM
2.5
species and PM
0.1
. Similar to ALAN, there may be elements of the built environment that are
being captured, with those living in cities having higher exposure to AP. There was moderate correlation
between ALAN and the Harvard AP model (Spearman correlation = 0.60, Table 4.5). The risk association
between both AP and NHL and ALAN and NHL were attenuated in a combined multivariate model (Table
4.6). Though not statistically significant, the protective direction of the effect seen in FL is particularly
interesting. FL is a more indolent disease and therefore may go unnoticed. There could be bias in
93

diagnosis in certain rural communities, but we did not see an association with rural census blocks. The
multivariate model, adjusted for SES, did attenuate this effect.
Overall, there appears to be modest risk of NHL for those living in areas of higher PM
2.5
, with a
statistically significant increased risk for those exposed to levels greater than the federal guidelines.

94

Table 4.1 Demographic characteristics of the California Teachers Study Cohort in air pollution analysis

Cohort NHL

n=100,861 n=762

PM
2.5
(μg/m
3
; mean, sd) 12.3 3.4 12.38 3.43
Participants residing in areas of PM
2.5
> 12 μg/m
3*
48,481 48.10% 388 50.90%

Race

White 86,265 85.5% 680 89.2%
Black 2,955 2.9% 15 2.0%
Hispanic 4,687 4.6% 20 2.6%
Asian/Pacific islander 4,753 4.7% 32 4.2%
Other 2,201 2.2% 15 2.0%

SES

Quartile 1 3,914 3.9% 14 1.9%
Quartile 2 15,733 15.8% 97 12.8%
Quartile 3 32,074 32.1% 237 31.4%
Quartile 4 48,146 48.2% 407 53.9%

Rural Census Block

Rural 10,179 10.2% 63 8.3%
Suburban 20,153 20.2% 132 17.5%
Urban 69,577 69.6% 560 74.2%

BMI

<20 10,613 11.0% 64 8.8%
20-24.9 48,617 50.3% 347 47.7%
25-29.9 23,923 24.7% 207 28.5%
30+ 13,529 14.0% 109 15.0%

Smoking

Never 66,953 66.8% 479 63.2%
Former 28,279 28.2% 246 32.5%
Current 4,964 5.0% 33 4.4%

Alcohol

Never 32,135 33.7% 239 32.8%
Former 55,541 58.2% 435 59.8%
Current 7,704 8.1% 54 7.4%

Any Daily NSAID

95

No 69,746 70.2% 523 69.8%
Yes 29,569 29.8% 226 30.2%

Exercise

Did not meet AHA guidelines 34,816 34.8% 237 31.5%
Met AHA guidelines 65,317 65.2% 516 68.5%

Family History of NHL

No 92,387 95.2% 675 91.5%
Yes 4,640 4.8% 63 8.5%

Any exposure to pesticides

No 63,318 62.8% 538 70.6%
Yes 37,543 37.2% 224 29.4%

Histology

DLBCL

189 24.8%
FL

138 18.1%
CLL/SLL

177 23.2%
Marginal Zone

85 11.2%
Other

173 22.7%

* Average exposure across follow-up
96

Table 4.2 Association between particulate matter and non-Hodgkin lymphoma

Overall NHL (n=762) DLBCL (n=189) FL (n=138) CLL/SLL (n=177)
Age adjusted HR 95% CI HR 95% CI HR 95% CI HR 95% CI
per 10 μg/m
3
PM
2.5

1.06 (0.89-1.27) 1.07 (0.75-1.53) 0.87 (0.58-1.31) 1.08 (0.75-1.57)
PM
2.5
> 12 μg/m
3

1.19 (1.03-1.38) 1.27 (0.94-1.71) 0.95 (0.68-1.33) 1.22 (0.90-1.65)

Rural Census Block
Rural 1.00 - 1.00 - 1.00 - 1.00 -
Suburban 1.06 (0.78-1.43) 1.01 (0.55-1.84) 1.09 (0.55-2.16) 0.98 (0.51-1.86)
Urban 1.21 (0.94-1.57) 1.19 (0.71-2.00) 1.15 (0.63-2.10) 1.30 (0.75-2.25)

Census Block SES

Quartile 1 1.00 - 1.00 - 1.00 - NA NA
Quartile 2 1.74 (0.99-3.05) 1.53 (0.45-5.20) 2.83 (0.67-12.0) 1.00 -
Quartile 3 2.03 (1.18-3.47) 2.41 (0.76-7.70) 2.19 (0.54-9.07) 1.26 (0.77-2.09)
Quartile 4 2.04 (1.20-3.48) 2.53 (0.80-7.98) 2.54 (0.62-10.4) 1.29 (0.80-2.07)

BMI

<20 0.98 (0.75-1.28) 1.23 (0.75-2.04) 0.90 (0.46-1.75) 0.98 (0.58-1.68)
20-24.9 1.00 - 1.00 - 1.00 - 1.00 -
25-29.9 1.08 (0.91-1.28) 1.11 (0.78-1.58) 1.28 (0.85-1.92) 0.90 (0.63-1.29)
30+ 1.09 (0.88-1.36) 1.40 (0.93-2.10) 1.55 (0.97-2.48) 0.66 (0.39-1.12)

Smoking

Never 1.00 - 1.00 - 1.00 - 1.00 -
Former 0.99 (0.85-1.16) 0.96 (0.70-1.32) 0.84 (0.58-1.23) 1.01 (0.74-1.39)
Current 0.87 (0.61-1.23) 1.04 (0.55-1.99) 0.70 (0.28-1.71) 0.33 (0.10-1.03)

Alcohol

Never 1.00 - 1.00 - 1.00 - 1.00 -
97

Former 1.01 (0.86-1.18) 0.98 (0.72-1.35) 0.98 (0.67-1.42) 1.26 (0.90-1.760
Current 0.79 (0.89-1.07) 0.52 (0.26-1.05) 1.18 (0.65-2.160 0.91 (0.49-1.68)

Any Daily NSAID

No 1.00 - 1.00 - 1.00 - 1.00 -
Yes 0.94 (0.80-1.10) 0.71 (0.51-0.99) 1.08 (0.75-1.54) 1.05 (0.76-1.44)

Exercise

Did not meet AHA guidelines 1.00 - 1.00 - 1.00 - 1.00 -
Met AHA guidelines 1.14 (0.98-1.33) 1.07 (0.79-1.46) 1.49 (1.02-2.18) 1.00 (0.73-1.37)

Family History of NHL

No 1.00 - 1.00 - 1.00 - 1.00 -
Yes 1.12 (0.96-1.31) 1.03 (0.74-1.44) 1.34 (0.98-1.84) 1.01 (0.71-1.43)

Any exposure to pesticides

No 1.00 - 1.00 - 1.00 - 1.00 -
Yes 0.89 (0.76-1.05) 0.79 (0.57-1.11) 1.16 (0.80-1.67) 0.68 (0.47-0.97)

Overall NHL (n=762) DLBCL (n=189) FL (n=138) CLL/SLL (n=177)
Multivariate* HR 95% CI HR 95% CI HR 95% CI HR 95% CI
per 10 μg/m
3
PM
2.5
1.11 (0.93-1.33) 1.14 (0.80-1.64) 0.93 (0.62-1.41) 1.12 (0.77-1.63)
PM
2.5
> 12 μg/m
3
1.22 (1.05-1.41) 1.32 (0.98-1.77) 0.97 (0.69-1.37) 1.22 (0.90-1.66)
*Adjusted for age, race, SES quartile, and smoking status (never, former, current)

98

Table 4.3 Comparison of Kleeman and Harvard air pollution models

Cohort NHL

n=100,861 n=762

PM
0.1
Kleeman (μg/m
3
; mean, sd) 0.64 0.25 0.74 0.29
PM
2.5
Kleeman (μg/m
3
; mean, sd) 7.73 1.89 8.06 2.18
PM
2.5
Harvard (μg/m
3
; mean, sd) 12.2 3.5 12.6 4.08
PM
2.5
Kleeman metals (μg/m
3
; mean, sd) 0.27 0.09 0.28 0.09
PM
2.5
Kleeman organic carbons (μg/m
3
; mean, sd) 1.34 0.59 1.47 0.74

99

Table 4.4 Association of Kleeman air pollution model and non-Hodgkin lymphoma
Age adjusted Overall NHL (n=762) DLBCL (n=189) FL (n=138) CLL/SLL (n=177)

HR 95% CI HR 95% CI HR 95% CI HR 95% CI
PM
0.1
0.94 (0.72-1.22) 0.90 (0.53-1.53) 0.67 (0.33-1.35) 0.88 (0.50-1.54)
per 10 μg/m
3
PM
2.5

0.97 (0.69-1.35) 0.84 (0.43-1.64) 0.46 (0.21-0.99) 1.45 (0.73-2.86)
per 100 ng/m
3
PM
2.5
metals 0.96 (0.89-1.04) 0.92 (0.78-1.07) 0.74 (0.61-0.89) 1.09 (0.94-1.28)
PM
2.5
organic carbons 1.01 (0.91-1.11) 1.01 (0.83-1.24) 0.75 (0.58-0.97) 1.10 (0.91-1.35)

Multivariate* Overall NHL (n=762) DLBCL (n=189) FL (n=138) CLL/SLL (n=177)

HR 95% CI HR 95% CI HR 95% CI HR 95% CI
PM
0.1
0.98 (0.76-1.27) 0.92 (0.53-1.58) 0.75 (0.38-1.49) 0.88 (0.49-1.58)
per 10 μg/m
3
PM
2.5

1.04 (0.75-1.46) 0.92 (0.47-1.80) 0.50 (0.23-1.10) 1.62 (0.81-3.24)
per 100 ng/m
3
PM
2.5
metals 0.97 (0.90-1.05) 0.93 (0.79-1.08) 0.75 (0.62-0.90) 1.11 (0.95-1.30)
PM
2.5
organic carbons 1.02 (0.92-1.13) 1.01 (0.83-1.24) 0.77 (0.59-1.00) 1.13 (0.93-1.38)

* Adjusted for age, race, SES, and smoking status (never, former, current).

100

Table 4.5 Spearman correlation of particulate matter, World Atlas, and ultraviolet radiation

PM
2.5

Kleeman
PM
0.1

Kleeman
PM
2.5

Kleeman
metals
PM
2.5
Kleeman
organic carbons
World
Atlas
Ultraviolet
radiation
PM
2.5
Harvard
0.53 0.27 0.22 0.35 0.6 0.53
PM
2.5
Kleeman

0.53 0.66 0.84 0.53 0.26
PM
0.1
Kleeman

0.33 0.64 0.45 -0.15
PM
2.5
Kleeman
metals

0.59 0.24 0.08
PM
2.5
Kleeman
organic carbons

0.57 -0.03
World Atlas

0.09

101

Table 4.6 Multivariate model of light at night and air pollution and non-Hodgkin lymphoma risk in the California Teachers Study
Multivariate Model* Overall NHL (n=762) DLBCL (n=189) FL (n=138) CLL/SLL (n=177)

HR 95% CI HR 95% CI HR 95% CI HR 95% CI
per 10 μg/m
3
PM
2.5

1.07 (0.85-1.36) 1.01 (0.63-1.620 1.16 (0.68-1.97) 1.07 (0.65-1.76)
Light at Night

Quintile 1 1.00 - 1.00 - 1.00 - 1.00 -
Quintile 2 0.99 (0.77-1.25) 1.23 (0.75-2.03) 0.89 (0.53-1.50) 0.80 (0.48-1.32)
Quintile 3 1.04 (0.82-1.33) 1.30 (0.79-2.16) 0.81 (0.47-1.40) 0.84 (0.50-1.39)
Quintile 4 0.90 (0.69-1.18) 1.19 (0.69-2.04) 0.57 (0.30-1.07) 0.89 (0.52-1.52)
Quintile 5 1.13 (0.84-1.51) 1.35 (0.74-2.46) 0.76 (0.39-1.48) 1.04 (0.57-1.89)

*Adjusted for age, race, and SES

102

Figure 4.1 Exclusion/inclusion criteria for California Teachers Study participants in analysis of air
pollution and cancer

103

Figure 4.2 Change in Mean Concentration and Standard Deviation of PM2.5 (μg/m3) in the California
Teachers Study Cohort from 2000-2016

104

Figure 4.3 Distribution of PM2.5 in the California Teachers Study cohort

Mean PM
2.5
(μg/m
3
)
Standard Deviation PM
2.5
(μg/m
3
)
Mean PM
2.5
(μg/m
3
)
Standard Deviation PM
2.5
(μg/m
3
)
105

Figure 4.4 Distribution of PM0.1 in the California Teachers Study cohort

Mean PM
0.1
(μg/m
3
)
Standard Deviation PM
0.1
(μg/m
3
)
106

Chapter 5. Conclusion
The study of chronic inflammatory conditions and disease risk is difficult given the long duration
and multifaceted nature of the exposure. Through this dissertation, we have looked at just three forms
of inflammation. The complexity of the HLA region and huge variety of haplotypes makes identifying
specific associations difficult. While host immune factors from HLA are constant, environmental
exposures pose greater difficulty in assessing. Not only is there temporal variation, but typically we only
account for exposure at participants’ residence. While this may work for an exposure such as light at
night, it does introduce bias to other exposures such as air pollution. Use of geospatial measures of
exposure is becoming more prominent as technology improves and spatial resolution improves. As seen
in chapter 3, there has been an evolution in how we measure light at night; an exposure that has only
been studied the past few years. As our tools improved, we have gone from neighborhood level
measures (DMSP-OLS 2.7km) to block level measures (VIIRS/World Atlas 750m). The next potential
evolution is the use of photographs from the International Space Station. Through a more concerted
effort, it may be possible to improve our resolution on a global scale to 30 meters, residence level; and
even include data on wavelength of light. Machine learning tools and use of Google StreetView images
may also provide an avenue to assess building structure to better understand the amount of ambient
light that enters into residences.
Advances have been made as well in air pollution measurement. In chapter 4, we used various
emissions source based model and accounted for weather patterns to attempt at a more accurate
measure of particulate matter. While this improves upon use of single monitoring stations, we are
always going to be limited by our data. For instance, the Kleeman model makes use of a wildfire PM
inventory as this is an acute exposure that is unique to California; and therefore, not included in the
national Harvard model. However, we see that if the model is inaccurate, our results will be biased as
well. Different PM species and sizes act differently, making modeling a challenge. There is currently
107

work being done on satellite based measures which would allow for more direct measurement of PM
and it’s species (Diner et al., 2018).
The largest difficulty in use of spatial measurements are that they are inherently made outside
the home, and do not account for variation of the exposure inside the residence, an issue particularly
relevant to air pollution. We also do not account for exposure away from the home, such as at work.
The advantage of using these measures in a cohort setting is that one would expect these biases to be
non-differential in nature, therefore attenuating any effects. However, this means that a much larger
sample size would be needed to detect an association, which can be challenging for rarer conditions
such as non-Hodgkin lymphoma.
Each chapter only focused on a single source of inflammation. Inflammation is a complex
exposure that encompasses not just genetics and environment, but also individual behaviors such as
physical activity and diet. Several dietary indices of inflammation (Hébert, Shivappa, Wirth, Hussey, &
Hurley, 2019; Joanna, Holly, Håkan, Karl, & Alicja, 2018) have also been developed that could help form
a more complete picture of inflammation and disease risk. New large cohort studies, such as the All of
Us study are collecting these data, along with genetics to look at gene-environment interactions, though
it will be several years before outcomes will accrue.
Finally, it is necessary to address the heterogeneity of the outcome. NHL consists of dozens of
different diseases, each with differing etiologies. The 2014 InterLymph monograph provided the largest
study of its kind with an overview of risk factors by major NHL subtypes (Morton et al., 2014). In our
study, having obese BMI in the CTS seemed to be a risk factor for DLBCL and FL, yet protective for
CLL/SLL. These offsetting directions of risk can lead to not seeing an association for NHL overall. To
further complicate matters, recent molecular typing efforts; especially in DLBCL, have identified even
finer differentiation of subtypes. Identification and validation of the initial DLBCL subtypes of germinal
108

center and activated B-cell classification (Nowakowski & Czuczman, 2015; Rosenwald & Staudt, 2003)
has barely reached the clinic and we already have new methods that further separate DLBCL into four
possible subtypes (Schmitz et al., 2018). Replication of these studies would be warranted in larger
pooled efforts, such as the National Cancer Institute Cohort Consortium.
109

Bibliography

Abdou, A. M., Gao, X., Cozen, W., Cerhan, J. R., Rothman, N., Martin, M. P., . . . Wang, S. S. (2010).
Human leukocyte antigen (HLA) A1-B8-DR3 (8.1) haplotype, tumor necrosis factor (TNF) G-308A,
and risk of non-Hodgkin lymphoma. Leukemia, 24(5), 1055-1058. doi:10.1038/leu.2010.17
American Cancer Society. (2017). Cancer Facts & Figures 2017. Retrieved from
https://www.cancer.org/content/dam/cancer-org/research/cancer-facts-and-statistics/annual-
cancer-facts-and-figures/2017/cancer-facts-and-figures-2017.pdf
Akers, N. K., Curry, J. D., Conde, L., Bracci, P. M., Smith, M. T., & Skibola, C. F. (2011). Association of HLA-
DQB1 alleles with risk of follicular lymphoma. Leuk Lymphoma, 52(1), 53-58.
doi:10.3109/10428194.2010.532888
Al-Naggar, R. A., & Anil, S. (2016). Artificial Light at Night and Cancer: Global Study. Asian Pac J Cancer
Prev, 17(10), 4661-4664. doi:10.22034/APJCP.2016.17.10.4661
Alcoceba, M., Sebastián, E., Marín, L., Balanzategui, A., Sarasquete, M. E., Chillón, M. C., . . . González, M.
(2013). HLA specificities are related to development and prognosis of diffuse large B-cell
lymphoma. Blood, 122(8), 1448-1454. doi:10.1182/blood-2013-02-483420
Anderson, L. A., Atman, A. A., McShane, C. M., Titmarsh, G. J., Engels, E. A., & Koshiol, J. (2014). Common
infection-related conditions and risk of lymphoid malignancies in older individuals. Br J Cancer,
110(11), 2796-2803. doi:10.1038/bjc.2014.173
Bassig, B. A., Cerhan, J. R., Au, W. Y., Kim, H. N., Sangrajrang, S., Hu, W., . . . Rothman, N. (2015). Genetic
susceptibility to diffuse large B‐cell lymphoma in a pooled study of three Eastern Asian
populations. European Journal of Haematology, 95(5), 442-448. doi:doi:10.1111/ejh.12513
Baugh, K., Elvidge, C., Tilottama, G., & Ziskin, D. (2010). Development of a 2009 stable lights product
using DMSP-OLS data (Vol. 30).
Bell, M. L., Ebisu, K., Peng, R. D., & Dominici, F. (2009). Adverse health effects of particulate air pollution:
modification by air conditioning. Epidemiology (Cambridge, Mass.), 20(5), 682-686.
doi:10.1097/EDE.0b013e3181aba749
Benjamini, Y., & Hochberg, Y. (1995). Controlling the False Discovery Rate: A Practical and Powerful
Approach to Multiple Testing. Journal of the Royal Statistical Society. Series B (Methodological),
57(1), 289-300.
Berglund, M., Enblad, G., & Thunberg, U. (2011). SNP rs6457327 is a predictor for overall survival in
follicular lymphoma as well as survival after transformation. Blood, 118(16), 4489-4489.
doi:10.1182/blood-2011-06-362921
Berndt, S. I., Skibola, C. F., Joseph, V., Camp, N. J., Nieters, A., Wang, Z., . . . Slager, S. L. (2013). Genome-
wide association study identifies multiple risk loci for chronic lymphocytic leukemia. Nat Genet,
45(8), 868-876. doi:10.1038/ng.2652
http://www.nature.com/ng/journal/v45/n8/abs/ng.2652.html#supplementary-information
Bernstein, L., Allen, M., Anton-Culver, H., Deapen, D., & et al. (2002). High breast cancer incidence rates
among California teachers: Results from the California teachers study (United States). Cancer
Causes & Control, 13(7), 625-635. doi:Doi 10.1023/A:1019552126105
Bielska, M., Bojo, M., Klimkiewicz-Wojciechowska, G., Jesionek-Kupnicka, D., Borowiec, M., Kalinka-
Warzocha, E., . . . Lech-Maranda, E. (2015). Human leukocyte antigen-G polymorphisms
110

influence the clinical outcome in diffuse large B-cell lymphoma. Genes, Chromosomes and
Cancer, 54(3), 185-193. doi:doi:10.1002/gcc.22235
Blackwell, J. M., Jamieson, S. E., & Burgner, D. (2009). HLA and Infectious Diseases. Clinical Microbiology
Reviews, 22(2), 370-385. doi:10.1128/CMR.00048-08
Blask, D. E. (2009). Melatonin, sleep disturbance and cancer risk. Sleep Medicine Reviews, 13(4), 257-
264. doi:https://doi.org/10.1016/j.smrv.2008.07.007
Blask, D. E., Dauchy, R. T., Dauchy, E. M., Mao, L., Hill, S. M., Greene, M. W., . . . Davidson, L. (2014). Light
Exposure at Night Disrupts Host/Cancer Circadian Regulatory Dynamics: Impact on the Warburg
Effect, Lipid Signaling and Tumor Growth Prevention. PLoS ONE, 9(8), e102776.
doi:10.1371/journal.pone.0102776
Bortle, J. E. (2001). Introducing the Bortle dark-sky scale. Sky and Telescope, 101(2).
Boyle, T., Gallagher, R. P., Gascoyne, R. D., Connors, J. M., Le, N. D., & Spinelli, J. J. (2015). Lifetime
Physical Activity and the Risk of Non-Hodgkin Lymphoma. Cancer Epidemiology Biomarkers
& Prevention, 24(5), 873-877. doi:10.1158/1055-9965.epi-14-1303
Brunekreef, B., & Holgate, S. T. (2002). Air pollution and health. The Lancet, 360(9341), 1233-1242.
doi:https://doi.org/10.1016/S0140-6736(02)11274-8
Burchiel, S. W., & Luster, M. I. (2001). Signaling by Environmental Polycyclic Aromatic Hydrocarbons in
Human Lymphocytes. Clinical Immunology, 98(1), 2-10.
doi:https://doi.org/10.1006/clim.2000.4934
Cameron, P. U., Mallal, S. A., French, M. A. H., & Dawkins, R. L. (1990). Major histocompatibility complex
genes influence the outcome of HIV infection: Ancestral haplotypes with C4 null alleles explain
diverse HLA associations. Human Immunology, 29(4), 282-295.
doi:https://doi.org/10.1016/0198-8859(90)90042-N
Candore, G., Lio, D., Colonna Romano, G., & Caruso, C. (2002). Pathogenesis of autoimmune diseases
associated with 8.1 ancestral haplotype: effect of multiple gene interactions. Autoimmunity
Reviews, 1(1), 29-35. doi:https://doi.org/10.1016/S1568-9972(01)00004-0
Candore, G., Romano, G. C., D’Anna, C., Di Lorenzo, G., Gervasi, F., Lio, D., . . . Caruso, C. (1998).
Biological Basis of the HLA-B8,DR3-Associated Progression of Acquired Immune Deficiency
Syndrome. Pathobiology, 66(1), 33-37. doi:10.1159/000027992
Center for International Blood and Marrow Transplant, a contractor for the C.W. Bill Young Cell
Transplantation Program operated through the U. S. Department of Health and Human Services,
Health Resources and Services Administration, Healthcare Systems Bureau. U.S. Transplant Data
by Center Report, NHL - Non-Hodgkin Lymphoma, Number of Transplants Reported in 2016. U.S.
Transplant Data by Center Report. Last Updated: May 17, 2017. Retrieved from
https://bloodcell.transplant.hrsa.gov/research/transplant_data/us_tx_data/data_by_center/ce
nter.aspx
Cerhan, J. R., Berndt, S. I., Vijai, J., Ghesquieres, H., McKay, J., Wang, S. S., . . . Chanock, S. J. (2014).
Genome-wide association study identifies multiple susceptibility loci for diffuse large B cell
lymphoma. Nat Genet, 46(11), 1233-1238. doi:10.1038/ng.3105
http://www.nature.com/ng/journal/v46/n11/abs/ng.3105.html#supplementary-information
Cerhan, J. R., Fredericksen, Z. S., Novak, A. J., Ansell, S. M., Kay, N. E., Liebow, M., . . . Slager, S. L. (2012).
A two-stage evaluation of genetic variation in immune and inflammation genes with risk of non-
111

Hodgkin lymphoma identifies new susceptibility locus in 6p21.3 region. Cancer Epidemiol
Biomarkers Prev, 21(10), 1799-1806. doi:1055-9965.EPI-12-0696 [pii] 10.1158/1055-9965.EPI-
12-0696 [doi]
Cerhan, J. R., Fredericksen, Z. S., Wang, A. H., Habermann, T. M., Kay, N. E., Macon, W. R., . . . Liebow, M.
(2011). Design and validity of a clinic-based case-control study on the molecular epidemiology of
lymphoma. Int J Mol Epidemiol Genet, 2(2), 95-113.
Cerhan, J. R., Kricker, A., Paltiel, O., Flowers, C. R., Wang, S. S., Monnereau, A., . . . Skibola, C. F. (2014).
Medical History, Lifestyle, Family History, and Occupational Risk Factors for Diffuse Large B-Cell
Lymphoma: The InterLymph Non-Hodgkin Lymphoma Subtypes Project. JNCI Monographs,
2014(48), 15-25. doi:10.1093/jncimonographs/lgu010
Chang, E. T., Canchola, A. J., Clarke, C. A., Lu, Y., West, D. W., Bernstein, L., . . . Horn-Ross, P. L. (2011).
Dietary phytocompounds and risk of lymphoid malignancies in the California Teachers Study
cohort. Cancer Causes Control, 22(2), 237-249. doi:10.1007/s10552-010-9692-5
Chang, E. T., Canchola, A. J., Cockburn, M., Lu, Y., Wang, S. S., Bernstein, L., . . . Horn-Ross, P. L. (2011).
Adulthood residential ultraviolet radiation, sun sensitivity, dietary vitamin D, and risk of
lymphoid malignancies in the California Teachers Study. Blood, 118(6), 1591-1599.
doi:10.1182/blood-2011-02-336065
Choi, H.-B., Roh, S.-Y., Choi, E.-J., Yoon, H.-Y., Kim, S.-Y., Hong, Y.-S., . . . Kim, T.-G. (2008). Association of
HLA alleles with non-Hodgkin’s lymphoma in Korean population. International Journal of
Hematology, 87(2), 203-209. doi:10.1007/s12185-008-0040-4
Christiansen, F. T., Zhang, W. J., Griffiths, M., Mallal, S. A., & Dawkins, R. L. (1991). Major
histocompatibility complex (MHC) complement deficiency, ancestral haplotypes and systemic
lupus erythematosus (SLE): C4 deficiency explains some but not all of the influence of the MHC. J
Rheumatol, 18(9), 1350-1358.
Conde, L., Halperin, E., Akers, N. K., Brown, K. M., Smedby, K. E., Rothman, N., . . . Skibola, C. F. (2010).
Genome-wide association study of follicular lymphoma identifies a risk locus at 6p21.32. Nat
Genet, 42(8), 661-664. doi:ng.626 [pii] 10.1038/ng.626 [doi]
Conroy, S. M., Maskarinec, G., Morimoto, Y., Franke, A. A., Cooney, R. V., Wilkens, L. R., . . . Kolonel, L. N.
(2013). Non-Hodgkin Lymphoma and Circulating Markers of Inflammation and Adiposity in a
Nested Case–Control Study: The Multiethnic Cohort. Cancer Epidemiology Biomarkers &
Prevention, 22(3), 337-347. doi:10.1158/1055-9965.epi-12-0947
Crowther-Swanepoel, D., Broderick, P., Di Bernardo, M. C., Dobbins, S. E., Torres, M., Mansouri, M., . . .
Houlston, R. S. (2010). Common variants at 2q37.3, 8q24.21, 15q21.3 and 16q24.1 influence
chronic lymphocytic leukemia risk. Nat Genet, 42(2), 132-136. doi:10.1038/ng.510
Crowther-Swanepoel, D., Di Bernardo, M. C., Jamroziak, K., Karabon, L., Frydecka, I., Deaglio, S., . . .
Houlston, R. S. (2011). Common genetic variation at 15q25.2 impacts on chronic lymphocytic
leukaemia risk. British Journal of Haematology, 154(2), 229-233. doi:doi:10.1111/j.1365-
2141.2011.08706.x
Crowther-Swanepoel, D., Mansouri, M., Enjuanes, A., Vega, A., Smedby, K. E., Ruiz-Ponte, C., . . .
Houlston, R. S. (2010). Verification that common variation at 2q37.1, 6p25.3, 11q24.1, 15q23,
and 19q13.32 influences chronic lymphocytic leukaemia risk. British Journal of Haematology,
150(4), 473-479. doi:doi:10.1111/j.1365-2141.2010.08270.x
112

Darbre, P. D. (2018). Overview of air pollution and endocrine disorders. Int J Gen Med, 11, 191-207.
doi:10.2147/ijgm.s102230
de Bakker, P. I., McVean, G., Sabeti, P. C., Miretti, M. M., Green, T., Marchini, J., . . . Rioux, J. D. (2006). A
high-resolution HLA and SNP haplotype map for disease association studies in the extended
human MHC. Nat Genet, 38(10), 1166-1172. doi:10.1038/ng1885
De La Concha, E. G., Fernandez-Arquero, M., Martinez, A., Vidal, F., Vigil, P., Conejero, L., . . . Fontan, G.
(1999). HLA class II homozygosity confers susceptibility to common variable immunodeficiency
(CVID). Clinical and Experimental Immunology, 116(3), 516-520. doi:10.1046/j.1365-
2249.1999.00926.x
De Roos, A. J., Davis, S., Colt, J. S., Blair, A., Airola, M., Severson, R. K., . . . Ward, M. H. (2010). Residential
proximity to industrial facilities and risk of non-Hodgkin lymphoma. Environmental Research,
110(1), 70-78. doi:https://doi.org/10.1016/j.envres.2009.09.011
de Visser, K. E., & Coussens, L. M. (2006). The inflammatory tumor microenvironment and its impact on
cancer development. Contrib Microbiol, 13, 118-137. doi:10.1159/000092969
DellaValle, C. T., Deziel, N. C., Jones, R. R., Colt, J. S., De Roos, A. J., Cerhan, J. R., . . . Ward, M. H. (2016).
Polycyclic aromatic hydrocarbons: determinants of residential carpet dust levels and risk of non-
Hodgkin lymphoma. Cancer causes & control : CCC, 27(1), 1-13. doi:10.1007/s10552-015-0660-y
Di Bernardo, M. C., Broderick, P., Harris, S., Dyer, M. J. S., Matutes, E., Dearden, C., . . . Houlston, R. S.
(2013). Risk of developing chronic lymphocytic leukemia is influenced by HLA-A class I variation.
Leukemia, 27(1), 255-258.
doi:http://www.nature.com/leu/journal/v27/n1/suppinfo/leu2012173s1.html
Di Bernardo, M. C., Crowther-Swanepoel, D., Broderick, P., Webb, E., Sellick, G., Wild, R., . . . Houlston, R.
S. (2008). A genome-wide association study identifies six susceptibility loci for chronic
lymphocytic leukemia. Nat Genet, 40(10), 1204-1210.
doi:http://www.nature.com/ng/journal/v40/n10/suppinfo/ng.219_S1.html
Di, Q., Kloog, I., Koutrakis, P., Lyapustin, A., Wang, Y., & Schwartz, J. (2016). Assessing PM2.5 Exposures
with High Spatiotemporal Resolution across the Continental United States. Environmental
Science & Technology, 50(9), 4712-4721. doi:10.1021/acs.est.5b06121
Diner, D. J., Boland, S. W., Brauer, M., Bruegge, C., Burke, K. A., Chipman, R., . . . Xu, F. (2018). Advances
in multiangle satellite remote sensing of speciated airborne particulate matter and association
with adverse health effects: from MISR to MAIA (Vol. 12): SPIE.
Donaldson, K., Brown, D., Clouter, A., Duffin, R., MacNee, W., Renwick, L., . . . Stone, V. (2002). The
pulmonary toxicology of ultrafine particles. Journal of Aerosol Medicine-Deposition Clearance
and Effects in the Lung, 15(2), 213-220.
Donaldson, K., Stone, V., Clouter, A., Renwick, L., & MacNee, W. (2001). Ultrafine particles. Occupational
and Environmental Medicine, 58(3), 211-+. doi:Doi 10.1136/Oem.58.3.211
Dorak, M. T., Machulla, H. K. G., Hentschel, M., Mills, K. I., Langner, J., & Burnett, A. K. (1996). Influence
of the major histocompatibility complex on age at onset of chronic lymphoid leukaemia.
International Journal of Cancer, 65(2), 134-139. doi:doi:10.1002/(SICI)1097-
0215(19960117)65:2<134::AID-IJC2>3.0.CO;2-#
Edlefsen, K. L., Martínez-Maza, O., Madeleine, M. M., Magpantay, L., Mirick, D. K., Kopecky, K. J., . . . De
Roos, A. J. (2014). Cytokines in serum in relation to future non-Hodgkin lymphoma risk: Evidence
113

for associations by histologic subtype. International Journal of Cancer, 135(4), 913-922.
doi:10.1002/ijc.28724
El Ansary, M. M., Mohammed, L. A., Hassan, T. H., Baraka, A., & Ahmed, A. A. (2015). Human leukocyte
antigen-DRB1 polymorphism in childhood acute lymphoblastic leukemia. Molecular and Clinical
Oncology, 3(2), 425-429. doi:10.3892/mco.2014.466
Elvidge, C. D., Baugh, K., Zhizhin, M., Hsu, F. C., & Ghosh, T. (2017). VIIRS night-time lights. International
Journal of Remote Sensing, 38(21), 5860-5879. doi:10.1080/01431161.2017.1342050
Etter, J. L., Cannioto, R., Soh, K. T., Alquassim, E., Almohanna, H., Dunbar, Z., . . . Moysich, K. B. (2018).
Lifetime physical inactivity is associated with increased risk for Hodgkin and non-Hodgkin
lymphoma: A case-control study. Leukemia Research, 69, 7-11.
doi:https://doi.org/10.1016/j.leukres.2018.03.014
Falchi, F., Cinzano, P., Duriscoe, D., Kyba, C. C. M., Elvidge, C. D., Baugh, K., . . . Furgoni, R. (2016). The
new world atlas of artificial night sky brightness. Science Advances, 2(6), e1600377.
doi:10.1126/sciadv.1600377
Favero, G., Franceschetti, L., Bonomini, F., Rodella, L. F., & Rezzani, R. (2017). Melatonin as an Anti-
Inflammatory Agent Modulating Inflammasome Activation. International Journal of
Endocrinology, 2017, 13. doi:10.1155/2017/1835195
Foo, Jia N., Smedby, Karin E., Akers, Nicholas K., Berglund, M., Irwan, Ishak D., Jia, X., . . . Liu, J. (2013).
Coding Variants at Hexa-allelic Amino Acid 13 of HLA-DRB1 Explain Independent SNP
Associations with Follicular Lymphoma Risk. The American Journal of Human Genetics, 93(1),
167-172. doi:http://dx.doi.org/10.1016/j.ajhg.2013.05.020
García-Álvarez, M., Alcoceba, M., López-Parra, M., Puig, N., Antón, A., Balanzategui, A., . . . González, M.
(2017). HLA specificities are associated with prognosis in IGHV-mutated CLL-like high-count
monoclonal B cell lymphocytosis. PLoS ONE, 12(3), e0172978.
doi:10.1371/journal.pone.0172978
Garcia-Saenz, A., Miguel, A. S. d., Espinosa, A., Valentin, A., Aragonés, N., Llorca, J., . . . Kogevinas, M.
(2018). Evaluating the Association between Artificial Light-at-Night Exposure and Breast and
Prostate Cancer Risk in Spain (MCC-Spain Study). Environ Health Perspect, 126(4), 047011.
doi:doi:10.1289/EHP1837
Garcia, E., Berhane, K. T., Islam, T., McConnell, R., Urman, R., Chen, Z., & Gilliland, F. D. (2019).
Association of Changes in Air Quality With Incident Asthma in Children in California, 1993-
2014Association of Changes in Air Quality With Incident Asthma in Children in California, 1993-
2014Association of Changes in Air Quality With Incident Asthma in Children in California, 1993-
2014. JAMA, 321(19), 1906-1915. doi:10.1001/jama.2019.5357
Gatto, N. M., Deapen, D., Stoyanoff, S., Pinder, R., Narayan, S., Bordelon, Y., & Ritz, B. (2014). Lifetime
exposure to estrogens and Parkinson's disease in California teachers. Parkinsonism & Related
Disorders, 20(11), 1149-1156. doi:https://doi.org/10.1016/j.parkreldis.2014.08.003
Ghesquieres, H., Slager, S. L., Jardin, F., Veron, A. S., Asmann, Y. W., Maurer, M. J., . . . Cerhan, J. R.
(2015). Genome-Wide Association Study of Event-Free Survival in Diffuse Large B-Cell
Lymphoma Treated With Immunochemotherapy. J Clin Oncol, 33(33), 3930-3937.
doi:10.1200/jco.2014.60.2573
114

Giudice, A., Crispo, A., Grimaldi, M., Polo, A., Bimonte, S., Capunzo, M., . . . Montella, M. (2018). The
Effect of Light Exposure at Night (LAN) on Carcinogenesis via Decreased Nocturnal Melatonin
Synthesis. Molecules, 23(6), 1308.
Goel, N., Basner, M., Rao, H., & Dinges, D. F. (2013). Chapter Seven - Circadian Rhythms, Sleep
Deprivation, and Human Performance. In M. U. Gillette (Ed.), Progress in Molecular Biology and
Translational Science (Vol. 119, pp. 155-190): Academic Press.
González-Galarza, Faviel F., Takeshita, Louise Y. C., Santos, Eduardo J. M., Kempson, F., Maia, Maria
Helena T., Silva, Andrea Luciana Soares d., . . . Middleton, D. (2015). Allele frequency net 2015
update: new features for HLA epitopes, KIR and disease and HLA adverse drug reaction
associations. Nucleic Acids Research, 43(D1), D784-D788. doi:10.1093/nar/gku1166
Gourraud, P.-A., Lamiraux, P., El-Kadhi, N., Raffoux, C., & Cambon-Thomsen, A. (2005). Inferred HLA
Haplotype Information for Donors From Hematopoietic Stem Cells Donor Registries. Human
Immunology, 66(5), 563-570. doi:http://dx.doi.org/10.1016/j.humimm.2005.01.011
Gragert, L., Fingerson, S., Albrecht, M., Maiers, M., Kalaycio, M., & Hill, B. T. (2014). Fine-mapping of HLA
associations with chronic lymphocytic leukemia in US populations. Blood, 124(17), 2657-2665.
doi:10.1182/blood-2014-02-558767
Gragert, L., Madbouly, A., Freeman, J., & Maiers, M. (2013). Six-locus high resolution HLA haplotype
frequencies derived from mixed-resolution DNA typing for the entire US donor registry. Human
Immunology, 74(10), 1313-1320. doi:https://doi.org/10.1016/j.humimm.2013.06.025
Gruzieva, O., Merid, S. K., Gref, A., Gajulapuri, A., Lemonnier, N., Ballereau, S., . . . Pershagen, G. (2017).
Exposure to Traffic-Related Air Pollution and Serum Inflammatory Cytokines in Children. Environ
Health Perspect, 125(6), 067007. doi:10.1289/ehp460
Guðjónsson, J. E., Valdimarsson, H., Kárason, A., Antonsdóttir, A. A., Hjaltey Rúnarsdóttir, E., Gulcher, J.
R., & Stefánsson, K. (2002). HLA-Cw6-Positive and HLA-Cw6-Negative Patients with Psoriasis
Vulgaris have Distinct Clinical Features. Journal of Investigative Dermatology, 118(2), 362-365.
doi:http://dx.doi.org/10.1046/j.0022-202x.2001.01656.x
Haack, M., Sanchez, E., & Mullington, J. M. (2007). Elevated Inflammatory Markers in Response to
Prolonged Sleep Restriction Are Associated With Increased Pain Experience in Healthy
Volunteers. Sleep, 30(9), 1145-1152.
Habermann, T. M., Wang, S. S., Maurer, M. J., Morton, L. M., Lynch, C. F., Ansell, S. M., . . . Cerhan, J. R.
(2008). Host immune gene polymorphisms in combination with clinical and demographic factors
predict late survival in diffuse large B-cell lymphoma patients in the pre-rituximab era. Blood,
112(7), 2694-2702. doi:10.1182/blood-2007-09-111658
Hajat, A., Allison, M., Diez-Roux, A. V., Jenny, N. S., Jorgensen, N. W., Szpiro, A. A., . . . Kaufman, J. D.
(2015). Long-term exposure to air pollution and markers of inflammation, coagulation, and
endothelial activation: a repeat-measures analysis in the Multi-Ethnic Study of Atherosclerosis
(MESA). Epidemiology, 26(3), 310-320. doi:10.1097/ede.0000000000000267
Hanahan, D., & Weinberg, Robert A. (2011). Hallmarks of Cancer: The Next Generation. Cell, 144(5), 646-
674. doi:https://doi.org/10.1016/j.cell.2011.02.013
Hanoun, M., Eisele, L., Suzuki, M., Greally, J. M., Hüttmann, A., Aydin, S., . . . Dürig, J. (2012). Epigenetic
Silencing of the Circadian Clock Gene CRY1 is Associated with an Indolent Clinical Course in
Chronic Lymphocytic Leukemia. PLoS ONE, 7(3), e34347. doi:10.1371/journal.pone.0034347
115

Hébert, J. R., Shivappa, N., Wirth, M. D., Hussey, J. R., & Hurley, T. G. (2019). Perspective: The Dietary
Inflammatory Index (DII)—Lessons Learned, Improvements Made, and Future Directions.
Advances in Nutrition, 10(2), 185-195. doi:10.1093/advances/nmy071
Hogan, M. K., Kovalycsik, T., Sun, Q., Rajagopalan, S., & Nelson, R. J. (2015). Combined effects of
exposure to dim light at night and fine particulate matter on C3H/HeNHsd mice. Behavioural
Brain Research, 294(Supplement C), 81-88. doi:https://doi.org/10.1016/j.bbr.2015.07.033
Hojjat-Farsangi, M., Jeddi-Tehrani, M., Amirzargar, A. A., Razavi, S. M., Sharifian, R. A., Rabbani, H., &
Shokri, F. (2008). Human leukocyte antigen class II allele association to disease progression in
Iranian patients with chronic lymphocytic leukemia. Human Immunology, 69(10), 666-674.
doi:http://dx.doi.org/10.1016/j.humimm.2008.07.015
Hojjat-Farsangi, M., Razavi, S. M., Sharifian, R. A., & Shokri, F. (2014). Frequency analysis of HLA class I
alleles in Iranian patients with progressive and non-progressive chronic lymphocytic leukemia.
Human Immunology, 75(2), 170-175. doi:https://doi.org/10.1016/j.humimm.2013.11.003
Hojman, P., Gehl, J., Christensen, J. F., & Pedersen, B. K. (2018). Molecular Mechanisms Linking Exercise
to Cancer Prevention and Treatment. Cell Metabolism, 27(1), 10-21.
doi:10.1016/j.cmet.2017.09.015
Holly, E. A., Gautam, M., & Bracci, P. M. Comparison of Interviewed and Non-Interviewed Non-Hodgkin's
Lymphoma (NHL) Patients in the San Francisco Bay Area. Annals of Epidemiology, 12(6), 419-
425. doi:10.1016/S1047-2797(01)00287-3
Hosnijeh, F. S., Portengen, L., Spath, F., Bergdahl, I. A., Melin, B., Mattiello, A., . . . Vermeulen, R. (2016).
Soluble B-cell activation marker of sCD27 and sCD30 and future risk of B-cell lymphomas: A
nested case-control study and meta-analyses. Int J Cancer, 138(10), 2357-2367.
doi:10.1002/ijc.29969
Hu, J., Zhang, H., Chen, S., Ying, Q., Wiedinmyer, C., Vandenberghe, F., & Kleeman, M. J. (2014a).
Identifying PM and PM Sources for Epidemiological Studies in California. Environ Sci Technol,
48(9), 4980-4990. doi:10.1021/es404810z
Hu, J., Zhang, H., Chen, S., Ying, Q., Wiedinmyer, F., Vandenberghe, F., & Kleeman, M. J. (2014b).
Predicting Primary PM2.5 and PM0.1 Trace Composition for Epidemiological Studies in
California. Environmental Science and Technology, 48(9), 4971-4979.
Hu, J., Zhang, H., Ying, Q., Chen, S. H., Vandenberghe, F., & Kleeman, M. J. (2014). Long-term particulate
matter modeling for health effects studies in California – Part 1: Model performance on
temporal and spatial variations. Atmos. Chem. Phys. Discuss., 14(14), 20997-21036.
doi:10.5194/acpd-14-20997-2014
Hu, Z. H., Connett, J. E., Yuan, J. M., & Anderson, K. E. (2016). Role of survivor bias in pancreatic cancer
case-control studies. Ann Epidemiol, 26(1), 50-56. doi:10.1016/j.annepidem.2015.11.001
Hung, L.-J., Chan, T.-F., Wu, C.-H., Chiu, H.-F., & Yang, C.-Y. (2012). Traffic Air Pollution and Risk of Death
from Ovarian Cancer in Taiwan: Fine Particulate Matter (PM2.5) as a Proxy Marker. Journal of
Toxicology and Environmental Health, Part A, 75(3), 174-182.
doi:10.1080/15287394.2012.641200
Hurley, S., Goldberg, D., Nelson, D., Hertz, A., Horn-Ross, P. L., Bernstein, L., & Reynolds, P. (2014). Light
at Night and Breast Cancer Risk Among California Teachers. Epidemiology, 25(5), 697-706.
doi:10.1097/ede.0000000000000137
116

Huss, A., Wel, L. v., Bogaards, L., Vrijkotte, T., Wolf, L., Hoek, G., & Vermeulen, R. (2019). Shedding Some
Light in the Dark—A Comparison of Personal Measurements with Satellite-Based
Estimates of Exposure to Light at Night among Children in the Netherlands. Environ Health
Perspect, 127(6), 067001. doi:doi:10.1289/EHP3431
Hussain, S. K., Makgoeng, S. B., Everly, M. J., Goodman, M. T., Martinez-Maza, O., Morton, L. M., . . .
Engels, E. A. (2016). HLA and Risk of Diffuse Large B cell Lymphoma After Solid Organ
Transplantation. Transplantation, 100(11), 2453-2460. doi:10.1097/tp.0000000000001025
Hysing, M., Pallesen, S., Stormark, K. M., Jakobsen, R., Lundervold, A. J., & Sivertsen, B. (2015). Sleep and
use of electronic devices in adolescence: results from a large population-based study. BMJ
Open, 5(1), e006748. doi:10.1136/bmjopen-2014-006748
Irwin, M. R. (2015). Why Sleep Is Important for Health: A Psychoneuroimmunology Perspective. Annual
Review of Psychology, 66(1), 143-172. doi:10.1146/annurev-psych-010213-115205
James, P., Bertrand, K. A., Hart, J. E., Schernhammer, E. S., Tamimi, R. M., & Laden, F. (2017). Outdoor
Light at Night and Breast Cancer Incidence in the Nurses' Health Study II. Environ Health
Perspect, 125(8), 087010. doi:10.1289/ehp935
Joanna, K., Holly, H., Håkan, M., Karl, M., & Alicja, W. (2018). Questionnaire-Based Anti-Inflammatory
Diet Index as a Predictor of Low-Grade Systemic Inflammation. Antioxidants & Redox Signaling,
28(1), 78-84. doi:10.1089/ars.2017.7330
Johnson, P. C. D., McAulay, K. A., Montgomery, D., Lake, A., Shield, L., Gallagher, A., . . . Jarrett, R. F.
(2015). Modeling HLA associations with EBV-positive and -negative Hodgkin lymphoma suggests
distinct mechanisms in disease pathogenesis. International Journal of Cancer, 137(5), 1066-
1075. doi:10.1002/ijc.29467
Juszczynski, P., Kalinka, E., Bienvenu, J., Woszczek, G., Borowiec, M., Robak, T., . . . Warzocha, K. (2002).
Human leukocyte antigens class II and tumor necrosis factor genetic polymorphisms are
independent predictors of non-Hodgkin lymphoma outcome. Blood, 100(8), 3037-3040.
doi:10.1182/blood-2002-02-0654
Karabon, L., Jedynak, A., Giebel, S., Wolowiec, D., Kielbinski, M., Woszczyk, D., . . . Frydecka, I. (2011).
KIR/HLA gene combinations influence susceptibility to B-cell chronic lymphocytic leukemia and
the clinical course of disease. Tissue Antigens, 78(2), 129-138. doi:10.1111/j.1399-
0039.2011.01721.x
Kelly, F. J., & Fussell, J. C. (2015). Air pollution and public health: emerging hazards and improved
understanding of risk. Environmental Geochemistry and Health, 37(4), 631-649.
doi:10.1007/s10653-015-9720-1
Khan, R. K., & Strand, M. A. (2018). Road dust and its effect on human health: a literature review.
Epidemiol Health, 40(0), e2018013-2018010. doi:10.4178/epih.e2018013
Kim, K. Y., Lee, E., Kim, Y. J., & Kim, J. (2017). The association between artificial light at night and
prostate cancer in Gwangju City and South Jeolla Province of South Korea. Chronobiology
International, 34(2), 203-211. doi:10.1080/07420528.2016.1259241
Klein, S. L., Jedlicka, A., & Pekosz, A. (2010). The Xs and Y of immune responses to viral vaccines. The
Lancet Infectious Diseases, 10(5), 338-349. doi:http://dx.doi.org/10.1016/S1473-
3099(10)70049-9
117

Klitz, W., Gragert, L., Maiers, M., Fernandez-Viña, M., Ben-Naeh, Y., Benedek, G., . . . Israel, S. (2010).
Genetic differentiation of Jewish populations. Tissue Antigens, 76(6), 442-458.
doi:doi:10.1111/j.1399-0039.2010.01549.x
Koelwyn, G. J., Wennerberg, E., Demaria, S., & Jones, L. W. (2015). Exercise in Regulation of
Inflammation-Immune Axis Function in Cancer Initiation and Progression. Oncology (Williston
Park, N.Y.), 29(12), 908-922.
Koo, Y. S., Song, J.-Y., Joo, E.-Y., Lee, H.-J., Lee, E., Lee, S.-k., & Jung, K.-Y. (2016). Outdoor artificial light
at night, obesity, and sleep health: Cross-sectional analysis in the KoGES study. Chronobiology
International, 33(3), 301-314. doi:10.3109/07420528.2016.1143480
Koshiol, J., Lam, T. K., Gridley, G., Check, D., Brown, L. M., & Landgren, O. (2011). Racial Differences in
Chronic Immune Stimulatory Conditions and Risk of Non-Hodgkin's Lymphoma in Veterans From
the United States. Journal of Clinical Oncology, 29(4), 378-385. doi:10.1200/jco.2010.30.1515
Kovacevic, A., Mavros, Y., Heisz, J. J., & Fiatarone Singh, M. A. (2018). The effect of resistance exercise on
sleep: A systematic review of randomized controlled trials. Sleep Medicine Reviews, 39, 52-68.
doi:https://doi.org/10.1016/j.smrv.2017.07.002
Kreyling, W. G., Semmler, M., & Moller, W. (2004). Dosimetry and toxicology of ultrafine particles.
Journal of Aerosol Medicine-Deposition Clearance and Effects in the Lung, 17(2), 140-152.
doi:Doi 10.1089/0894268041457147
Kyba, C. C. M., & Aronson, K. J. (2015). Assessing Exposure to Outdoor Lighting and Health Risks.
Epidemiology, 26(4), e50. doi:10.1097/ede.0000000000000307
Kyba, C. C. M., Kuester, T., Sánchez de Miguel, A., Baugh, K., Jechow, A., Hölker, F., . . . Guanter, L.
(2017). Artificially lit surface of Earth at night increasing in radiance and extent. Science
Advances, 3(11). doi:10.1126/sciadv.1701528
Lahti, T. A., Partonen, T., Kyyrönen, P., Kauppinen, T., & Pukkala, E. (2008). Night-time work predisposes
to non-Hodgkin lymphoma. International Journal of Cancer, 123(9), 2148-2151.
doi:10.1002/ijc.23566
Lech-Maranda, E., Juszczynski, P., Szmigielska-Kaplon, A., Jamroziak, K., Balcerczak, E., & Robak, T.
(2007). Human leukocyte antigens HLA DRB1 influence clinical outcome of chronic lymphocytic
leukemia? Haematologica, 92(5), 710-711.
Lewy, A., Wehr, T., Goodwin, F., Newsome, D., & Markey, S. (1980). Light suppresses melatonin
secretion in humans. Science, 210(4475), 1267-1269. doi:10.1126/science.7434030
Li, W., Dorans, K. S., Wilker, E. H., Rice, M. B., Ljungman, P. L., Schwartz, J. D., . . . Mittleman, M. A.
(2017). Short-Term Exposure to Ambient Air Pollution and Biomarkers of Systemic Inflammation.
The Framingham Heart Study, 37(9), 1793-1800. doi:10.1161/atvbaha.117.309799
Lim, U., Kocarnik, J. M., Bush, W. S., Matise, T. C., Caberto, C., Park, S. L., . . . Le Marchand, L. (2014).
Pleiotropy of Cancer Susceptibility Variants on the Risk of Non-Hodgkin Lymphoma: The PAGE
Consortium. PLoS ONE, 9(3), e89791. doi:10.1371/journal.pone.0089791
Linet, M. S., Vajdic, C. M., Morton, L. M., de Roos, A. J., Skibola, C. F., Boffetta, P., . . . Chiu, B. C. H.
(2014). Medical History, Lifestyle, Family History, and Occupational Risk Factors for Follicular
Lymphoma: The InterLymph Non-Hodgkin Lymphoma Subtypes Project. JNCI Monographs,
2014(48), 26-40. doi:10.1093/jncimonographs/lgu006
118

Lipsett, M. J., Ostro, B. D., Reynolds, P., Goldberg, D., Hertz, A., Jerrett, M., . . . Bernstein, L. (2011). Long-
term exposure to air pollution and cardiorespiratory disease in the California teachers study
cohort. Am J Respir Crit Care Med, 184(7), 828-835. doi:10.1164/rccm.201012-2082OC
Liukkonen, T., Räsänen, P., Ruokonen, A., Laitinen, J., Jokelainen, J., Leinonen, M., . . . Timonen, M.
(2007). C-reactive protein Levels and Sleep Disturbances: Observations Based on The Northern
Finland 1966 Birth Cohort Study. Psychosomatic Medicine, 69(8), 756-761.
doi:10.1097/PSY.0b013e318157cb96
Loomis, D., Grosse, Y., Lauby-Secretan, B., Ghissassi, F. E., Bouvard, V., Benbrahim-Tallaa, L., . . . Straif, K.
(2013). The carcinogenicity of outdoor air pollution. The Lancet Oncology, 14(13), 1262-1263.
doi:10.1016/s1470-2045(13)70487-x
Lu, Y., Abdou, A. M., Cerhan, J. R., Morton, L. M., Severson, R. K., Davis, S., . . . Wang, S. S. (2011). Human
leukocyte antigen class I and II alleles and overall survival in diffuse large B-cell lymphoma and
follicular lymphoma. ScientificWorldJournal, 11, 2062-2070. doi:10.1100/2011/373876 [doi]
Lu, Y., Prescott, J., Sullivan-Halley, J., Henderson, K. D., Ma, H., Chang, E. T., . . . Bernstein, L. (2009). Body
Size, Recreational Physical Activity, and B-Cell Non-Hodgkin Lymphoma Risk Among Women in
the California Teachers Study. American Journal of Epidemiology, 170(10), 1231-1240.
doi:10.1093/aje/kwp268
Lurmann, F., Avol, E., & Gilliland, F. (2015). Emissions reduction policies and recent trends in Southern
California’s ambient air quality. Journal of the Air & Waste Management Association, 65(3), 324-
335. doi:10.1080/10962247.2014.991856
Lyapustin, A., Martonchik, J., Wang, Y., Laszlo, I., & Korkin, S. (2011). Multiangle implementation of
atmospheric correction (MAIAC): 1. Radiative transfer basis and look-up tables. Journal of
Geophysical Research: Atmospheres, 116(D3). doi:10.1029/2010jd014985
Machulla, H. K. G., Muller, L. P., Schaaf, A., Kujat, G., Schunermarck, U., & Langner, J. (2001). Association
of chronic lymphocytic leukemia with specific alleles of the HLA-DR4:DR53:DQ8 haplotype in
German patients. International Journal of Cancer, 92(2), 203-207. doi:10.1002/1097-
0215(200102)9999:9999<::aid-ijc1167>3.0.co;2-a
Madbouly, A., Gragert, L., Freeman, J., Leahy, N., Gourraud, P. A., Hollenbach, J. A., . . . Maiers, M.
(2014). Validation of statistical imputation of allele-level multilocus phased genotypes from
ambiguous HLA assignments. Tissue Antigens, 84(3), 285-292. doi:10.1111/tan.12390
Magnani, C., Ranucci, A., Badaloni, C., Cesaroni, G., Ferrante, D., Miligi, L., . . . Forastiere, F. (2016). Road
Traffic Pollution and Childhood Leukemia: A Nationwide Case-control Study in Italy. Archives of
Medical Research, 47(8), 694-705. doi:https://doi.org/10.1016/j.arcmed.2017.02.001
Makgoeng, S. B., Bolanos, R. S., Jeon, C. Y., Weiss, R. E., Arah, O. A., Breen, E. C., . . . Hussain, S. K. (2018).
Markers of Immune Activation and Inflammation, and Non-Hodgkin Lymphoma: A Meta-Analysis
of Prospective Studies. JNCI cancer spectrum, 2(4), pky082-pky082. doi:10.1093/jncics/pky082
Manor, S., Halagan, M., Shriki, N., Yaniv, I., Zisser, B., Maiers, M., . . . Stein, J. (2016). High-resolution HLA
A ∼B ∼DRB1 haplotype frequencies from the Ezer Mizion Bone Marrow Donor Registry in Israel.
Human Immunology, 77(12), 1114-1119. doi:https://doi.org/10.1016/j.humimm.2016.09.004
Mastrangelo, G., Fadda, E., & Marzia, V. (1996). Polycyclic aromatic hydrocarbons and cancer in man.
Environ Health Perspect, 104(11), 1166-1170. doi:doi:10.1289/ehp.961041166
119

McAulay, K. A., & Jarrett, R. F. (2015). Human leukocyte antigens and genetic susceptibility to
lymphoma. Tissue Antigens, 86(2), 98-113. doi:10.1111/tan.12604
McDonald, B. C., Goldstein, A. H., & Harley, R. A. (2015). Long-Term Trends in California Mobile Source
Emissions and Ambient Concentrations of Black Carbon and Organic Aerosol. Environmental
Science & Technology, 49(8), 5178-5188. doi:10.1021/es505912b
Meier-Ewert, H. K., Ridker, P. M., Rifai, N., Regan, M. M., Price, N. J., Dinges, D. F., & Mullington, J. M.
(2004). Effect of sleep loss on C-Reactive protein, an inflammatory marker of cardiovascular risk.
Journal of the American College of Cardiology, 43(4), 678-683.
doi:https://doi.org/10.1016/j.jacc.2003.07.050
Miller, M. A., Kandala, N.-B., Kivimaki, M., Kumari, M., Brunner, E. J., Lowe, G. D. O., . . . Cappuccio, F. P.
(2009). Gender Differences in the Cross-Sectional Relationships Between Sleep Duration and
Markers of Inflammation: Whitehall II Study. Sleep, 32(7), 857-864.
Montes-Ares, O., Moya-Quiles, M. R., Montes-Casado, M., Guerra-Pérez, N., Campillo, J. A., González, C.,
. . . Álvarez-López, M. R. (2006). Human leucocyte antigen-C in B chronic lymphocytic leukaemia.
British Journal of Haematology, 135(4), 517-519. doi:10.1111/j.1365-2141.2006.06334.x
Moore, S. C., Lee, I.-M., Weiderpass, E., Campbell, P. T., Sampson, J. N., Kitahara, C. M., . . . Patel, A. V.
(2016). Association of Leisure-Time Physical Activity With Risk of 26 Types of Cancer in 1.44
Million AdultsLeisure-Time Physical Activity and Risk of 26 Types of Cancer Leisure-Time Physical
Activity and Risk of 26 Types of Cancer. JAMA Internal Medicine, 176(6), 816-825.
doi:10.1001/jamainternmed.2016.1548
Mortazavi, S. A. R., Parhoodeh, S., Hosseini, M. A., Arabi, H., Malakooti, H., Nematollahi, S., . . .
Mortazavi, S. M. J. (2018). Blocking Short-Wavelength Component of the Visible Light Emitted by
Smartphones' Screens Improves Human Sleep Quality. Journal of biomedical physics &
engineering, 8(4), 375-380.
Morton, L. M., Slager, S. L., Cerhan, J. R., Wang, S. S., Vajdic, C. M., Skibola, C. F., . . . Sampson, J. N.
(2014). Etiologic Heterogeneity Among Non-Hodgkin Lymphoma Subtypes: The InterLymph Non-
Hodgkin Lymphoma Subtypes Project. JNCI Monographs, 2014(48), 130-144.
doi:10.1093/jncimonographs/lgu013
Morton, L. M., Turner, J. J., Cerhan, J. R., Linet, M. S., Treseler, P. A., Clarke, C. A., . . . Weisenburger, D.
D. (2007). Proposed classification of lymphoid neoplasms for epidemiologic research from the
Pathology Working Group of the International Lymphoma Epidemiology Consortium
(InterLymph). Blood, 110(2), 695-708. doi:10.1182/blood-2006-11-051672
Mueller, L. P., & Machulla, H. K. G. (2002). Increased Frequency of Homozygosity for HLA Class II Loci in
Female Patients with Chronic Lymphocytic Leukemia. Leukemia & Lymphoma, 43(5), 1013-1019.
doi:10.1080/10428190290021588
Nabavi, S. M., Nabavi, S. F., Sureda, A., Xiao, J., Dehpour, A. R., Shirooie, S., . . . Daglia, M. (2018). Anti-
inflammatory effects of Melatonin: A mechanistic review. Critical Reviews in Food Science and
Nutrition, 1-13. doi:10.1080/10408398.2018.1487927
Naugler, C., & Liwski, R. (2010). Human leukocyte antigen class I alleles and the risk of chronic
myelogenous leukemia: a meta-analysis. Leukemia & Lymphoma, 51(7), 1288-1292.
doi:doi:10.3109/10428191003802340
120

Nielsen, K. R., Steffensen, R., Bendtsen, M. D., Rodrigo-Domingo, M., Baech, J., Haunstrup, T. M., . . .
Johnsen, H. E. (2015). Inherited Inflammatory Response Genes Are Associated with B-Cell Non-
Hodgkin’s Lymphoma Risk and Survival. PLoS ONE, 10(10), e0139329.
doi:10.1371/journal.pone.0139329
Niens, M., Jarrett, R. F., Hepkema, B., Nolte, I. M., Diepstra, A., Platteel, M., . . . van den Berg, A. (2007).
HLA-A*02 is associated with a reduced risk and HLA-A*01 with an increased risk of developing
EBV+ Hodgkin lymphoma. Blood, 110(9), 3310-3315. doi:10.1182/blood-2007-05-086934
Nieters, A., Conde, L., Slager, S. L., Brooks-Wilson, A., Morton, L., Skibola, D. R., . . . Skibola, C. F. (2012).
PRRC2A and BCL2L11 gene variants influence risk of non-Hodgkin lymphoma: results from the
InterLymph consortium. Blood, 120(23), 4645-4648. doi:blood-2012-05-427989 [pii]
10.1182/blood-2012-05-427989 [doi]
Noble, J. A. (2015). Immunogenetics of type 1 diabetes: A comprehensive review. Journal of
Autoimmunity, 64, 101-112. doi:https://doi.org/10.1016/j.jaut.2015.07.014
Nowak, J., Kalinka-Warzocha, E., Juszczynski, P., Bilinski, P., Mika-Witkowska, R., Zajko, M., . . .
Warzocha, K. (2007). Association of human leukocyte antigen ancestral haplotype 8.1 with
adverse outcome of non-Hodgkin's lymphoma. Genes Chromosomes Cancer, 46(5), 500-507.
doi:10.1002/gcc.20436
Nowak, J., Kalinka-Warzocha, E., Juszczyński, P., Mika-Witkowska, R., Zajko, M., Graczyk-Pol, E., . . .
Warzocha, K. (2008). Haplotype-specific pattern of association of human major
histocompatibility complex with non-Hodgkin’s lymphoma outcome. Tissue Antigens, 71(1), 16-
26. doi:10.1111/j.1399-0039.2007.00954.x
Nowakowski, G. S., & Czuczman, M. S. (2015). ABC, GCB, and Double-Hit Diffuse Large B-Cell Lymphoma:
Does Subtype Make a Difference in Therapy Selection? Am Soc Clin Oncol Educ Book, e449-457.
doi:10.14694/EdBook_AM.2015.35.e449
Oberdorster, G. (2000). Toxicology of ultrafine particles: in vivo studies. Philosophical Transactions of
the Royal Society of London Series a-Mathematical Physical and Engineering Sciences,
358(1775), 2719-2739. doi:DOI 10.1098/rsta.2000.0680
Orouji, E., Tavakkol Afshari, J., Badiee, Z., Shirdel, A., & Alipour, A. (2012). Association between HLA-
DQB1 gene and patients with acute lymphoblastic leukemia (ALL). International Journal of
Hematology, 95(5), 551-555. doi:10.1007/s12185-012-1051-8
Ostro, B., Hu, J., Goldberg, D., Reynolds, P., Hertz, A., Bernstein, L., & Kleeman, M. J. (2015). Associations
of mortality with long-term exposures to fine and ultrafine particles, species and sources: results
from the California Teachers Study Cohort. Environ Health Perspect, 123(6), 549-556.
doi:10.1289/ehp.1408565
Ostro, B., Lipsett, M., Reynolds, P., Goldberg, D., Hertz, A., Garcia, C., . . . Bernstein, L. (2010). Long-term
exposure to constituents of fine particulate air pollution and mortality: results from the
California Teachers Study. Environ Health Perspect, 118(3), 363-369. doi:10.1289/ehp.0901181
Parodi, S., Santi, I., Casella, C., Puppo, A., Montanaro, F., Fontana, V., . . . Stagnaro, E. (2015). Risk of
leukaemia and residential exposure to air pollution in an industrial area in Northern Italy: a case-
control study. Int J Environ Health Res, 25(4), 393-404. doi:10.1080/09603123.2014.958136
Parodi, S., Santi, I., Marani, E., Casella, C., Puppo, A., Vercelli, M., & Stagnaro, E. (2014). Risk of Non-
Hodgkin's Lymphoma and Residential Exposure to Air Pollution in an Industrial Area in Northern
121

Italy: A Case-Control Study. Archives of Environmental & Occupational Health, 69(3), 139-147.
doi:10.1080/19338244.2013.763756
Paunić, V., Gragert, L., Madbouly, A., Freeman, J., & Maiers, M. (2012). Measuring Ambiguity in HLA
Typing Methods. PLoS ONE, 7(8), e43585. doi:10.1371/journal.pone.0043585
Pociot, F., Briant, L., Jongeneel, C. V., Mölvig, J., Worsaae, H., Abbal, M., . . . Cambon-Thomsen, A.
(1993). Association of tumor necrosis factor (TNF) and class II major histocompatibility complex
alleles with the secretion of TNF-a and TNF-0 by human mononuclear cells: a possible link to
insulin-dependent diabetes mellitus. European Journal of Immunology, 23(1), 224-231.
doi:doi:10.1002/eji.1830230135
Psaltopoulou, T., Sergentanis, T. N., Ntanasis-Stathopoulos, I., Tzanninis, I.-G., Riza, E., & Dimopoulos, M.
A. (2019). Anthropometric characteristics, physical activity and risk of hematological
malignancies: A systematic review and meta-analysis of cohort studies. International Journal of
Cancer, 145(2), 347-359. doi:10.1002/ijc.32109
Pun, V. C., Kazemiparkouhi, F., Manjourides, J., & Suh, H. H. (2017). Long-Term PM2.5 Exposure and
Respiratory, Cancer, and Cardiovascular Mortality in Older US Adults. American Journal of
Epidemiology, 186(8), 961-969. doi:10.1093/aje/kwx166
Qian, J., Block, G. D., Colwell, C. S., & Matveyenko, A. V. (2013). Consequences of Exposure to Light at
Night on the Pancreatic Islet Circadian Clock and Function in Rats. Diabetes, 62(10), 3469-3478.
doi:10.2337/db12-1543
Qiao, Y., Zhou, Y., Wu, C., Zhai, K., Han, X., Chen, J., . . . Lin, D. (2013). Risk of genome-wide association
study-identified genetic variants for non-Hodgkin lymphoma in a Chinese population.
Carcinogenesis, 34(7), 1516-1519. doi:10.1093/carcin/bgt082
Rahib, L., Smith, B. D., Aizenberg, R., Rosenzweig, A. B., Fleshman, J. M., & Matrisian, L. M. (2014).
Projecting Cancer Incidence and Deaths to 2030: The Unexpected Burden of Thyroid, Liver, and
Pancreas Cancers in the United States. Cancer Research, 74(11), 2913-2921. doi:10.1158/0008-
5472.can-14-0155
Ramis, R., Vidal, E., García-Pérez, J., Lope, V., Aragonés, N., Pérez-Gómez, B., . . . López-Abente, G.
(2009). Study of non-Hodgkin's lymphoma mortality associated with industrial pollution in Spain,
using Poisson models. BMC Public Health, 9(1), 26. doi:10.1186/1471-2458-9-26
Ratcliffe, M., Burd, C., Holder, K., & Fields, A. (2016). Defining Rural at the U.S. Census Bureau.
Retrieved from https://www2.census.gov/geo/pdfs/reference/ua/Defining_Rural.pdf
Rea, M. S., Brons, J. A., & Figueiro, M. G. (2011). Measurements of Light at Night (LAN) for a Sample of
Female School Teachers. Chronobiology International, 28(8), 673-680.
doi:10.3109/07420528.2011.602198
Reiter, R. J., Tan, D.-X., Korkmaz, A., Erren, T. C., Piekarski, C., Tamura, H., & Manchester, L. C. (2007).
Light at Night, Chronodisruption, Melatonin Suppression, and Cancer Risk: A Review. 13(4), 303-
328. doi:10.1615/CritRevOncog.v13.i4.30
Riemersma, S. A., Jordanova, E. S., Haasnoot, G. W., Drabbels, J., Schuuring, E., Schreuder, G. M. T., &
Kluin, P. M. (2006). The Relationship Between HLA Class II Polymorphisms and Somatic Deletions
in Testicular B Cell Lymphomas of Dutch Patients. Human Immunology, 67(4–5), 303-310.
doi:http://dx.doi.org/10.1016/j.humimm.2006.03.019
122

Robinson, J., Halliwell, J. A., Hayhurst, J. D., Flicek, P., Parham, P., & Marsh, Steven G E. (2015). The IPD
and IMGT/HLA database: allele variant databases. Nucleic Acids Research, 43(Database issue),
D423-D431. doi:10.1093/nar/gku1161
Rosenwald, A., & Staudt, L. M. (2003). Gene expression profiling of diffuse large B-cell lymphoma. Leuk
Lymphoma, 44 Suppl 3, S41-47.
Rothman, N., Skibola, C. F., Wang, S. S., Morgan, G., Lan, Q., Smith, M. T., . . . Nieters, A. (2006). Genetic
variation in TNF and IL10 and risk of non-Hodgkin lymphoma: a report from the InterLymph
Consortium. Lancet Oncol, 7(1), 27-38. doi:S1470-2045(05)70434-4 [pii] 10.1016/S1470-
2045(05)70434-4 [doi]
Sauvet, F., Leftheriotis, G., Gomez-Merino, D., Langrume, C., Drogou, C., Van Beers, P., . . . Chennaoui,
M. (2010). Effect of acute sleep deprivation on vascular function in healthy subjects. Journal of
Applied Physiology, 108(1), 68-75. doi:10.1152/japplphysiol.00851.2009
Schmitz, R., Wright, G. W., Huang, D. W., Johnson, C. A., Phelan, J. D., Wang, J. Q., . . . Staudt, L. M.
(2018). Genetics and Pathogenesis of Diffuse Large B-Cell Lymphoma. New England Journal of
Medicine, 378(15), 1396-1407. doi:10.1056/NEJMoa1801445
Seidemann, K., Zimmermann, M., Book, M., Meyer, U., Burkhardt, B., Welte, K., . . . Stanulla, M. (2005).
Tumor necrosis factor and lymphotoxin alfa genetic polymorphisms and outcome in pediatric
patients with non-Hodgkin's lymphoma: results from Berlin-Frankfurt-Munster Trial NHL-BFM
95. J Clin Oncol, 23(33), 8414-8421. doi:10.1200/jco.2005.01.2179
Shaflik, C. (1997). Environmental Effects of Roadway Lighting.
Shearer, W. T., Reuben, J. M., Mullington, J. M., Price, N. J., Lee, B.-N., Smith, E. O. B., . . . Dinges, D. F.
(2001). Soluble TNF-α receptor 1 and IL-6 plasma levels in humans subjected to the sleep
deprivation model of spaceflight. Journal of Allergy and Clinical Immunology, 107(1), 165-170.
doi:https://doi.org/10.1067/mai.2001.112270
Sillé, F. C. M., Conde, L., Zhang, J., Akers, N. K., Sanchez, S., Maltbaek, J., . . . Skibola, C. F. (2013).
Follicular lymphoma-protective HLA class II variants correlate with increased HLA-DQB1 protein
expression. Genes And Immunity, 15, 133. doi:10.1038/gene.2013.64
https://www.nature.com/articles/gene201364#supplementary-information
Singh, A., Kamal, R., Ahamed, I., Wagh, M., Bihari, V., Sathian, B., & Kesavachandran, C. N. (2018). PAH
exposure-associated lung cancer: an updated meta-analysis. Occupational Medicine, 68(4), 255-
261. doi:10.1093/occmed/kqy049
Skibola, C. F., Akers, N. K., Conde, L., Ladner, M., Hawbecker, S. K., Cohen, F., . . . Bracci, P. M. (2012).
Multi-locus HLA class I and II allele and haplotype associations with follicular lymphoma. Tissue
Antigens, 79(4), 279-286. doi:10.1111/j.1399-0039.2012.01845.x [doi]
Skibola, Christine F., Berndt, Sonja I., Vijai, J., Conde, L., Wang, Z., Yeager, M., . . . Rothman, N. (2014).
Genome-wide Association Study Identifies Five Susceptibility Loci for Follicular Lymphoma
outside the HLA Region. The American Journal of Human Genetics, 95(4), 462-471.
doi:http://dx.doi.org/10.1016/j.ajhg.2014.09.004
Skibola, C. F., Bracci, P. M., Halperin, E., Conde, L., Craig, D. W., Agana, L., . . . Brown, K. M. (2009).
Genetic variants at 6p21.33 are associated with susceptibility to follicular lymphoma. Nat Genet,
41(8), 873-875. doi:ng.419 [pii] 10.1038/ng.419 [doi]
123

Skibola, C. F., Bracci, P. M., Nieters, A., Brooks-Wilson, A., de Sanjose, S., Hughes, A. M., . . . Wang, S. S.
(2010). Tumor necrosis factor (TNF) and lymphotoxin-alpha (LTA) polymorphisms and risk of
non-Hodgkin lymphoma in the InterLymph Consortium. Am J Epidemiol, 171(3), 267-276.
doi:kwp383 [pii]10.1093/aje/kwp383 [doi]
Skibola, C. F., Conde, L., Foo, J. N., Riby, J., Humphreys, K., Sille, F. C., . . . Smedby, K. E. (2012). A meta-
analysis of genome-wide association studies of follicular lymphoma. BMC Genomics, 13, 516.
doi:10.1186/1471-2164-13-516
Slager, S. L., Camp, N. J., Conde, L., Shanafelt, T. D., Achenbach, S. J., Rabe, K. G., . . . Cerhan, J. R. (2012).
Common variants within 6p21.31 locus are associated with chronic lymphocytic leukaemia and,
potentially, other non-Hodgkin lymphoma subtypes. Br J Haematol, 159(5), 572-576.
doi:10.1111/bjh.12070 [doi]
Slager, S. L., Goldin, L. R., Strom, S. S., Lanasa, M. C., Spector, L. G., Rassenti, L., . . . Cerhan, J. R. (2010).
Genetic susceptibility variants for chronic lymphocytic leukemia. Cancer Epidemiol Biomarkers
Prev, 19(4), 1098-1102. doi:1055-9965.EPI-09-1217 [pii]10.1158/1055-9965.EPI-09-1217 [doi]
Slager, S. L., Rabe, K. G., Achenbach, S. J., Vachon, C. M., Goldin, L. R., Strom, S. S., . . . Cerhan, J. R.
(2011). Genome-wide association study identifies a novel susceptibility locus at 6p21.3 among
familial CLL. Blood, 117(6), 1911-1916. doi:10.1182/blood-2010-09-308205
Slager, S. L., Skibola, C. F., Di Bernardo, M. C., Conde, L., Broderick, P., McDonnell, S. K., . . . Houlston, R.
S. (2012). Common variation at 6p21.31 (BAK1) influences the risk of chronic lymphocytic
leukemia. Blood, 120(4), 843-846. doi:blood-2012-03-413591 [pii]10.1182/blood-2012-03-
413591 [doi]
Smedby, K. E., Foo, J. N., Skibola, C. F., Darabi, H., Conde, L., Hjalgrim, H., . . . Liu, J. (2011). GWAS of
follicular lymphoma reveals allelic heterogeneity at 6p21.32 and suggests shared genetic
susceptibility with diffuse large B-cell lymphoma. PLoS Genet, 7(4), e1001378.
doi:10.1371/journal.pgen.1001378 [doi]
Smedby, K. E., & Ponzoni, M. (2017). The aetiology of B‐cell lymphoid malignancies with a focus on
chronic inflammation and infections. Journal of Internal Medicine, 282(5), 360-370.
doi:doi:10.1111/joim.12684
Smigoc Schweiger, D., Mendez, A., Kunilo Jamnik, S., Bratanic, N., Bratina, N., Battelino, T., . . . Vidan-
Jeras, B. (2016). High-risk genotypes HLA-DR3-DQ2/DR3-DQ2 and DR3-DQ2/DR4-DQ8 in co-
occurrence of type 1 diabetes and celiac disease. Autoimmunity, 49(4), 240-247.
doi:10.3109/08916934.2016.1164144
Smith, M. T., Jones, R. M., & Smith, A. H. (2007). Benzene Exposure and Risk of Non-Hodgkin Lymphoma.
Cancer Epidemiology Biomarkers & Prevention, 16(3), 385-391. doi:10.1158/1055-
9965.epi-06-1057
Speedy, H. E., Di Bernardo, M. C., Sava, G. P., Dyer, M. J. S., Holroyd, A., Wang, Y., . . . Houlston, R. S.
(2014). A genome-wide association study identifies multiple susceptibility loci for chronic
lymphocytic leukemia. Nat Genet, 46(1), 56-60. doi:10.1038/ng.2843
http://www.nature.com/ng/journal/v46/n1/abs/ng.2843.html#supplementary-information
Straif, K., Baan, R., Grosse, Y., Secretan, B., Ghissassi, F. E., Bouvard, V., . . . Cogliano, V. (2007).
Carcinogenicity of shift-work, painting, and fire-fighting. The Lancet Oncology, 8(12), 1065-1066.
doi:10.1016/S1470-2045(07)70373-X
124

Sun, S.-C. (2017). The non-canonical NF-κB pathway in immunity and inflammation. Nature Reviews
Immunology, 17, 545. doi:10.1038/nri.2017.52
Swerdlow, S. H., Campo, E., Pileri, S. A., Harris, N. L., Stein, H., Siebert, R., . . . Jaffe, E. S. (2016). The 2016
revision of the World Health Organization classification of lymphoid neoplasms. Blood, 127(20),
2375-2390. doi:10.1182/blood-2016-01-643569
Tähkämö, L., Partonen, T., & Pesonen, A.-K. (2019). Systematic review of light exposure impact on
human circadian rhythm. Chronobiology International, 36(2), 151-170.
doi:10.1080/07420528.2018.1527773
Tan, D. E. K., Foo, J. N., Bei, J.-X., Chang, J., Peng, R., Zheng, X., . . . Liu, J. (2013). Genome-wide
association study of B cell non-Hodgkin lymphoma identifies 3q27 as a susceptibility locus in the
Chinese population. Nature Genetics, 45, 804. doi:10.1038/ng.2666
https://www.nature.com/articles/ng.2666#supplementary-information
Taylor, G. M., Hussain, A., Verhage, V., Thompson, P. D., Fergusson, W. D., Watkins, G., . . . Birch, J. M.
(2009). Strong association of the HLA-DP6 supertype with childhood leukaemia is due to a single
allele, DPB1[ast]0601. Leukemia, 23(5), 863-869.
Ten, L.-C., Chin, Y.-M., Tai, M.-C., Chin, E. F.-M., Lim, Y.-Y., Suthandiram, S., . . . Ng, C.-C. (2017). SNP
variants associated with non-Hodgkin lymphoma (NHL) correlate with human leukocyte antigen
(HLA) class II expression. Scientific Reports, 7, 41400. doi:10.1038/srep41400
Theodorou, I., Abel, L., Mauro, F., Duprey, B., Magnac, C., Payelle-Brogard, B., . . . Dighiero, G. (2002).
High occurence of DRB1 11 in chronic lymphocytic leukaemia families. British Journal of
Haematology, 119(3), 713-715. doi:doi:10.1046/j.1365-2141.2002.03926.x
Thio, C. L., Thomas, D. L., Karacki, P., Gao, X., Marti, D., Kaslow, R. A., . . . Carrington, M. (2003).
Comprehensive Analysis of Class I and Class II HLA Antigens and Chronic Hepatitis B Virus
Infection. Journal of Virology, 77(22), 12083-12087. doi:10.1128/jvi.77.22.12083-12087.2003
Tong, Z., Chen, Y., Malkawi, A., Adamkiewicz, G., & Spengler, J. D. (2016). Quantifying the impact of
traffic-related air pollution on the indoor air quality of a naturally ventilated building.
Environment International, 89-90, 138-146. doi:https://doi.org/10.1016/j.envint.2016.01.016
Troiano, R. P. (2018). Physical Activity Guidelines for Americans From the US Department of Health and
Human Services.
Turner, J. J., Morton, L. M., Linet, M. S., Clarke, C. A., Kadin, M. E., Vajdic, C. M., . . . Weisenburger, D. D.
(2010). InterLymph hierarchical classification of lymphoid neoplasms for epidemiologic research
based on the WHO classification (2008): update and future directions. Blood, 116(20), e90-e98.
doi:10.1182/blood-2010-06-289561
Turner, M. C., Krewski, D., Diver, W. R., Pope, C. A., 3rd, Burnett, R. T., Jerrett, M., . . . Gapstur, S. M.
(2017). Ambient Air Pollution and Cancer Mortality in the Cancer Prevention Study II. Environ
Health Perspect, 125(8), 087013. doi:10.1289/ehp1249
van Leeuwen, W. M. A., Lehto, M., Karisola, P., Lindholm, H., Luukkonen, R., Sallinen, M., . . . Alenius, H.
(2009). Sleep Restriction Increases the Risk of Developing Cardiovascular Diseases by
Augmenting Proinflammatory Responses through IL-17 and CRP. PLoS ONE, 4(2), e4589.
doi:10.1371/journal.pone.0004589
Vendrame, E., Hussain, S. K., Breen, E. C., Magpantay, L. I., Widney, D. P., Jacobson, L. P., . . . Martínez-
Maza, O. (2014). Serum Levels of Cytokines and Biomarkers for Inflammation and Immune
125

Activation, and HIV-Associated Non-Hodgkin B-Cell Lymphoma Risk. Cancer Epidemiology
Biomarkers & Prevention, 23(2), 343-349. doi:10.1158/1055-9965.epi-13-0714
Vendrame, E., & Martinez-Maza, O. (2011). Assessment of pre-diagnosis biomarkers of immune
activation and inflammation: insights on the etiology of lymphoma. J Proteome Res, 10(1), 113-
119. doi:10.1021/pr100729z
Vijai, J., Kirchhoff, T., Schrader, K. A., Brown, J., Dutra-Clarke, A. V., Manschreck, C., . . . Offit, K. (2013).
Susceptibility loci associated with specific and shared subtypes of lymphoid malignancies. PLoS
Genet, 9(1), e1003220. doi:10.1371/journal.pgen.1003220
Vijai, J., Wang, Z., Berndt, S. I., Skibola, C. F., Slager, S. L., de Sanjose, S., . . . Nieters, A. (2015). A
genome-wide association study of marginal zone lymphoma shows association to the HLA
region. Nat Commun, 6, 5751. doi:10.1038/ncomms6751
Wagner, B., da Silva Nardi, F., Schramm, S., Kraemer, T., Celik, A. A., Durig, J., . . . Rebmann, V. (2017).
HLA-E allelic genotype correlates with HLA-E plasma levels and predicts early progression in
chronic lymphocytic leukemia. Cancer, 123(5), 814-823. doi:10.1002/cncr.30427
Wang, S. S., Abdou, A. M., Morton, L. M., Thomas, R., Cerhan, J. R., Gao, X., . . . Carrington, M. (2010).
Human leukocyte antigen class I and II alleles in non-Hodgkin lymphoma etiology. Blood,
115(23), 4820-4823. doi:blood-2010-01-266775 [pii] 10.1182/blood-2010-01-266775 [doi]
Wang, S. S., Carrington, M., Berndt, S. I., Slager, S. L., Bracci, P. M., Voutsinas, J., . . . Skibola, C. F. (2018).
HLA Class I and II Diversity Contributes to the Etiologic Heterogeneity of Non-Hodgkin
Lymphoma Subtypes. Cancer Research, 78(14), 4086-4096. doi:10.1158/0008-5472.can-17-2900
Wang, S. S., Cozen, W., Cerhan, J. R., Colt, J. S., Morton, L. M., Engels, E. A., . . . Hartge, P. (2007).
Immune mechanisms in non-Hodgkin lymphoma: joint effects of the TNF G308A and IL10
T3575A polymorphisms with non-Hodgkin lymphoma risk factors. Cancer Res, 67(10), 5042-
5054. doi:67/10/5042 [pii] 10.1158/0008-5472.CAN-06-4752 [doi]
Wang, S. S., & Hartge, P. (2010). Etiology and epidemiology of non-hodgkin lymphoma. In J. O. Armitage,
P. M. Mauch, N. L. Harris, B. Coiffier, & D.-F. R. (Eds.), Non-Hodgkin Lymphomas (2 ed., pp. 64-
82). Philadelphia, PA, USA: Lippincott, Williams and Wilkens.
Wang, S. S., Lu, Y., Rothman, N., Abdou, A. M., Cerhan, J. R., De Roos, A., . . . Hartge, P. (2011). Variation
in effects of non-Hodgkin lymphoma risk factors according to the human leukocyte antigen
(HLA)-DRB1*01:01 allele and ancestral haplotype 8.1. PLoS ONE, 6(11), e26949.
doi:10.1371/journal.pone.0026949 [doi] PONE-D-11-12868 [pii]
Wang, S. S., Purdue, M. P., Cerhan, J. R., Zheng, T., Menashe, I., Armstrong, B. K., . . . Rothman, N. (2009).
Common gene variants in the tumor necrosis factor (TNF) and TNF receptor superfamilies and
NF-kB transcription factors and non-Hodgkin lymphoma risk. PLoS ONE, 4(4), e5360.
doi:10.1371/journal.pone.0005360 [doi]
Warzocha, K., Ribeiro, P., Bienvenu, J., Roy, P., Charlot, C., Rigal, D., . . . Salles, G. (1998). Genetic
Polymorphisms in the Tumor Necrosis Factor Locus Influence Non-Hodgkin's Lymphoma
Outcome. Blood, 91(10), 3574-3581.
Warzocha, K., Ribeiro, P., Renard, N., Bienvenu, J., Charlot, C., Coiffier, B., & Salles, G. (2000). Expression
of genes coding for the tumor necrosis factor and lymphotoxin ligand-receptor system in non-
Hodgkin's lymphomas. Cancer Immunology, Immunotherapy, 49(9), 469-475.
126

Warzocha, K., Salles, G., Bienvenu, J., Barbier, Y., Bastion, Y., Doche, C., . . . Coiffier, B. (1997). Prognostic
significance of TNF alpha and its p55 soluble receptor in malignant lymphomas. Leukemia, 11
Suppl 3, 441-443.
Warzocha, K., Salles, G., Bienvenu, J., Bastion, Y., Dumontet, C., Renard, N., . . . Coiffier, B. (1997). Tumor
necrosis factor ligand-receptor system can predict treatment outcome in lymphoma patients. J
Clin Oncol, 15(2), 499-508. doi:10.1200/jco.1997.15.2.499
Wilson, A. G., de Vries, N., Pociot, F., di Giovine, F. S., van der Putte, L. B., & Duff, G. W. (1993). An allelic
polymorphism within the human tumor necrosis factor alpha promoter region is strongly
associated with HLA A1, B8, and DR3 alleles. The Journal of Experimental Medicine, 177(2), 557-
560. doi:10.1084/jem.177.2.557
Wrench, D., Leighton, P., Skibola, C. F., Conde, L., Cazier, J. B., Matthews, J., . . . Fitzgibbon, J. (2011). SNP
rs6457327 in the HLA region on chromosome 6p is predictive of the transformation of follicular
lymphoma. Blood, 117(11), 3147-3150. doi:blood-2010-10-315382 [pii] 10.1182/blood-2010-10-
315382 [doi]
Wulff, K., Porcheret, K., Cussans, E., & Foster, R. G. (2009). Sleep and circadian rhythm disturbances:
multiple genes and multiple phenotypes. Current Opinion in Genetics & Development, 19(3),
237-246. doi:https://doi.org/10.1016/j.gde.2009.03.007
Xu, X., Jiang, S. Y., Wang, T.-Y., Bai, Y., Zhong, M., Wang, A., . . . Sun, Q. (2013). Inflammatory Response
to Fine Particulate Air Pollution Exposure: Neutrophil versus Monocyte. PLoS ONE, 8(8), e71414.
doi:10.1371/journal.pone.0071414
Xuecao, L., & Zhou, Y. (2017). Urban mapping using DMSP/OLS stable night-time light: a review.
Yeh, H.-L., Hsu, S.-W., Chang, Y.-C., Chan, T.-C., Tsou, H.-C., Chang, Y.-C., & Chiang, P.-H. (2017). Spatial
Analysis of Ambient PM(2.5) Exposure and Bladder Cancer Mortality in Taiwan. International
Journal of Environmental Research and Public Health, 14(5), 508. doi:10.3390/ijerph14050508
Younes, A., & Aggarwall, B. B. (2003). Clinical implications of the tumor necrosis factor family in benign
and malignant hematologic disorders. Cancer, 98(3), 458-467. doi:doi:10.1002/cncr.11524
Zhang, Z., Hong, Y., & Liu, N. (2017). Association of ambient Particulate matter 2.5 with intensive care
unit admission due to pneumonia: a distributed lag non-linear model. Scientific Reports, 7(1),
8679. doi:10.1038/s41598-017-08984-x

Abstract (if available)

Abstract Few risk factors have been identified in non-Hodgkin lymphoma (NHL) etiology, but as a cancer that originates in the immune cells, factors that severely alter the immune response have long been suspected. Immune dysregulation and chronic inflammation are recognized as a central factors in the etiology of the leading causes of morbidity and mortality, including cancers, cardiovascular diseases, and diabetes. Chronic inflammation, which reflects sustained activation of the innate and adaptive immune systems is recognized as one of the hallmarks of cancer and is recognized as a major risk factor for a growing number of cancers. In this dissertation, multiple sources of inflammation were examined in relation to non-Hodgkin lymphoma risk.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

Associations between exercise and quality of life in adult survivors of non-Hodgkin's lymphoma and colorectal cancer

PDF

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

PDF

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

PDF

Genetic risk factors in multiple myeloma

PDF

Assessing the impact of air pollution on adverse birth outcomes in a low resource setting

PDF

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

PDF

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

PDF

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

PDF

Study of risk factors of B-cell non-Hodgkin lymphoma (NHL) in the California Teachers Study

PDF

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

PDF

Genomic risk factors associated with Ewing Sarcoma susceptibility

PDF

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

PDF

Hormone therapy timing hypothesis and atherosclerosis

PDF

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

PDF

The effect of cytomegalovirus on gene expression of pediatric acute lymphoblastic leukemia

PDF

Age related macular degeneration in Latinos: risk factors and impact on quality of life

PDF

The role of steroid hormones in the etiology of urologic diseases

PDF

Reproductive factors and risk of B-cell non-Hodgkin's lymphomas among women in Los Angeles

PDF

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

Asset Metadata

Creator Zhong, Charlie (author)

Core Title The role of inflammation in non-Hodgkin lymphoma etiology

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Publication Date 09/18/2020

Defense Date 06/20/2019

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

Tag cohort,Epidemiology,non-Hodgkin lymphoma,OAI-PMH Harvest

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Franklin, Meredith (committee chair), Wiemels, Joseph (committee chair), Longcore, Travis (committee member), McKean-Cowdin, Roberta (committee member), Wang, Sophia (committee member)

Creator Email charlie.zhong@channing.harvard.edu,czhong@coh.org

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-218177

Unique identifier UC11674775

Identifier etd-ZhongCharl-7823.pdf (filename),usctheses-c89-218177 (legacy record id)

Legacy Identifier etd-ZhongCharl-7823.pdf

Dmrecord 218177

Document Type Dissertation

Rights Zhong, Charlie

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