Attachment 1 AsiaLymph Revised Protocol

A multi-center international hospital-based case-control study of lymphoma in Asia (AsiaLymph)

Co-Principal Investigators

Qing Lan (OEEB) and Nathaniel Rothman (OEEB): Overall study conduct

Lead Investigators

Martha Linet, Alina Brenner (REB): Study design

Lindsay Morton (REB), John Chan (Queen Elizabeth Hospital, Hong Kong): Pathology review and tumor molecular analyses

Charles Rabkin (IIB): Viral studies

Mark Purdue (OEEB): Occupational TCE exposure

Lynn Goldin, Neil Caporaso (GEB): Family history, CLL studies

Stephen Chanock (CGR, LTG): Genomics

Roel Vermeulen (Utrecht Univ., The Netherlands), Melissa Friesen (OEEB), and Mary Ward (OEEB): Occupational and environmental exposure assessment

Nilnajan Chatterjee, Kai Yu (BB): Study design and statistical analysis

Wei Hu (OEEB), Bryan Bassig (OEEB) and H. Dean Hosgood (Einstein Medical School): Field training, quality control, and study management at NCI

Jun Xu (Univ. of Hong Kong): Field study management in Asia

Summary

The contribution of environmental, occupational and genetic factors to lymphoma has generated a series of novel findings in studies of Caucasians. However, none of the chemical associations have been conclusively established and the identification of the key, functional alleles in gene regions associated with risk of NHL requires further elucidation. Further, the ability to follow-up, confirm, and extend these observations is limited by the low prevalence and limited range of several important chemical exposures and the high to complete linkage disequilibrium among key candidate genetic loci in Western populations. To optimize the ability to build on and clarify these findings, it is necessary to investigate populations that differ from Caucasians in both exposure patterns and underlying genetic structure. A multidisciplinary case-control study of lymphoma in Asia provides an opportunity to replicate and extend recent and novel observations made in studies among Caucasians in a population that is distinctly different with regard to patterns of key risk factors, including range of exposures, prevalence of exposures, correlations between exposures, and variation in gene regions of particular interest. Thus, a hospital-based case-control study of lymphoma in Eastern Asia (i.e., AsiaLymph) of 4,200 cases and 4,200 controls to be enrolled over a three-year period will be conducted. In addition, 2,000 cases of myeloid leukemias will be enrolled as well. The major postulated risk factors for evaluation in this study are chemical exposures (i.e., organochlorines, trichloroethylene, and benzene) and genetic susceptibility. Other factors potentially related to NHL, such as viral infections, UV exposure, medical conditions, and other lifestyle factors will also be explored. A particularly noteworthy aspect of AsiaLymph is central pathology review with immunophenotyping by two of the world’s leading lymphoma pathologists, which will enable accurate analysis of findings by molecular and histologic subtypes. AsiaLymph represents the optimal next step in the DCEG lymphoma portfolio. AsiaLymph should confirm and extend previous findings, and yield novel insights into the causes of lymphoma and leukemia in both Asia and the West.

Background and Rationale

Important leads have emerged from studies focused on the etiology of lymphoma, particularly those that include novel observations on the role of chemical exposures in the environmental and occupational setting and genetic susceptibility. These studies include the NCI-SEER non-Hodgkin’s Lymphoma (NHL) case-control study (Wang et al. 2006;De Roos et al. 2005;Colt et al. 2005;Colt et al. 2009;Stewart 2009; Purdue et al. 2010), a case-control study of NHL among women in Connecticut (Lan et al. 2006), a case-control study of NHL in New South Wales, Australia (Purdue et al. 2007), InterLymph pooled genetic studies (Rothman et al. 2006; Conde et al. 2010; Skibola et al. 2009), and studies of organochlorines in the US and Europe (De Roos et al. 2005;Engel et al. 2007;Rothman et al. 1997). This body of research suggests that organochlorines, trichloroethylene (TCE), and benzene may be associated with risk of lymphoma and that benzene and TCE have immunotoxic properties (Stewart 2009; Lan et al. 2004; Lan 2009), that genetic variation in certain loci involved in immunologic regulation (e.g., TNF/LTA, IL10, and IL4) may contribute to risk of lymphoma, and that interactions between these chemicals and genes may exist (Lan et al. 2004; Colt et al. 2009; Wang et al. 2007). However, none of these chemical or genetic associations have been conclusively established, and the underlying biologic plausibility, including identification of critical functional alleles in genetic studies, requires further elucidation. At the same time, there is a growing appreciation of the critical need for high quality pathology review in etiologic studies of lymphoma, as evidence is increasing that some risk factors may be highly specific to one or more subtypes of lymphoma (Morton et al. 2008). A multidisciplinary case-control study of lymphoma in Asia is timely because it will provide an opportunity to replicate and extend recent and novel observations made in studies among Caucasians in a population that is distinctly different with regard to patterns of key risk factors.

Lymphoma in Asia

Although NHL rates historically have been lower in Asia than in the West, there is evidence that rates have been rising in recent decades in Shanghai and Singapore (Jin et al. 1999; Chia et al. 2001). For example, in Shanghai between 1972-3 and 1993-4, NHL rates rose 33% in males and 66% in females, while there was a small drop in incidence rates for leukemia in both sexes. Overall, there was an 11% and 13% decline in the incidence rates of all cancers for males and females, respectively, during this time period (Jin et al. 1999). The distribution of NHL histologic subtypes also differs in Asians and Caucasians. Although diffuse large B-cell lymphoma (DLBCL) is the most common histologic subtype in both Asians and Caucasians, rates of follicular lymphoma are substantially lower in Asians, whereas rates of T-cell lymphomas, particularly nasal type NK/T-cell lymphomas, are substantially higher in Asians (Au and Lo 2005; Ng et al. 1986; Gross et al. 2008; Kadin et al. 1983). As a consequence, this study provides a unique opportunity to replicate and extend key findings observed in Caucasians for histologies with characteristics shared by both populations as well as to rigorously study the epidemiology of those tumors that appear to be more common in Asia than in Western populations.

Organochlorines

Organochlorine compounds (OCs) are chemicals comprised of joined carbon and chlorine atoms, and account for several classes of chemicals including dioxin, polychlorinated biphenyls (PCBs), and pentachlorophenols (PCPs). OCs are primarily synthetic chemicals that were first introduced in the 1940’s, and have been widely used as insulators and pesticides. OCs are relatively long-lasting, and can enter the environment through pesticide application, disposal of contaminated waste in landfills, and releases from manufacturing facilities that produce these chemicals (Centers for Disease Control, 2010). OCs have been suggested to be associated with a number of health concerns, including thyroid, metabolic, and reproductive disorders, in addition to several cancers, although results have been inconsistent (Toft et al. 2004; Langer 2010; Longnecker et. al. 1997; Gallagher et al. 2010; Purdue et al. 2009). While most OCs are banned, some are still being used in developing countries (e.g., DDT), and remain important environmental contaminants in the West.

A series of studies, many led by DCEG investigators over the last 12 years, have reported associations between NHL and plasma levels of several OCs, including DDT (rated by IARC as Group 2B, a possible carcinogen), PCBs (Group 2A, a probable carcinogen), and chlordane (Group 2B, a possible carcinogen), but it is not clear which class of compounds or specific congeners is primarily driving the association due in part to moderate to high correlation between them (Colt et al. 2005;De Roos et al. 2005;Engel et al. 2007;Rothman et al. 1997;Spinelli et al. 2007). In addition, only one study (the NCI-SEER study) has examined the relationship between plasma levels of more potent dioxin-like OCs and NHL, reporting associations for co-planar PCBs and dibenzofurans. At the same time, even though seven milliliters of plasma from each subject were used in that analysis, the most potent dioxin compounds could not be measured because they are present at very low levels (De Roos et al. 2005). Further, no epidemiologic study to date has been able to evaluate this finding because of the large volume of plasma required. Overall, the literature suggests that one or more components of OCs measured in blood, or some as yet unidentified factor associated with OCs, are causally related to risk of NHL, but the specific compounds have not been identified with confidence. It is thus important to disentangle the different specific OCs and associated factors to further our understanding of the role of environmental exposures in lymphomagenesis. Such efforts will also have important public health implications with regard to current use of DDT in developing countries and the need to carry out environmental clean-up of sites contaminated by OCs in the West [e.g., the Hudson River (Environmental Protection Agency, 2008)].

There are several advantages to studying OCs and lymphoma in Asia including much higher plasma levels and a wider range of several OCs (e.g., DDE, the major DDT metabolite) than in the West, differences in correlation among certain compounds, and substantial differences in plasma levels between countries in Asia due to different industrialization histories and pesticide use patterns (e.g., higher DDE and lower PCB levels in China vs. lower DDE and higher levels of other OCs in Taiwan) (Hsu et al. 2009;Lee et al. 2007; Gammon et al. 2002; Stellman et al.1998). As a consequence, the higher levels and wider range of exposure for several chemicals, the weaker overall correlation pattern between certain compounds, and the availability of a large plasma sample from study subjects to measure potent dioxin-like and dioxin chemicals will provide us with a unique opportunity to assess chemical-specific OC associations with NHL, which will complement previous and ongoing efforts to study these associations in the West.

Trichloroethylene (TCE)

TCE, a chlorinated solvent used in several industries primarily for metal degreasing, is one of the most important ground water contaminants in the US, and has been studied in cohort and case-control studies in the West. TCE is commonly found in ground water, and has been estimated to be present in about one-third of municipal water supplies in the United States (Jollow et al. 2009). While most water supplies are thought to have a relatively small concentration of TCE (i.e. parts per trillion or billion), samples taken from areas near direct contamination sites may have concentrations on the order of several hundred parts per million (ppm) (Jollow et al., 2009).

The carcinogenicity of TCE has been well studied with generally inconsistent results, although there is some suggestion that occupational rather than environmental exposures may be most relevant (Raaschou-Nielsen et al. 2003; Jollow et al. 2009). Cohort and case-control studies (including NCI-SEER) with extensive exposure assessment have generally observed associations with NHL, typically at relatively high estimated levels of exposure (Raaschou-Nielsen et al. 2003; Scott and Chiu 2006; Stewart 2009; Wartenberg et al. 2000; Purdue et al. 2010). Despite the research carried out to date, a connection between TCE and lymphoma has still not been established. It is currently rated by IARC as a probable carcinogen (Group 2A).

It is not feasible to conduct new cohort studies of TCE in the West or Asia due to current instability of the industrial workforce. Case-control studies of NHL in the US and Europe have been hampered by the low prevalence of exposure in these populations and the fact that most occupational exposures occurred in the distant past, as occupational use of TCE has been reduced and several manufacturing processes that use TCE have moved to developing countries (Mandel et al. 2006). In addition, collective interpretations of the various studies have been difficult due to differing exposure assessment methodologies and lack of evidence concerning exposure response trends (Mandel et al. 2006). In contrast, due to the extensive use of TCE in Asia, a higher proportion of the population is exposed, and there is a wide range of exposure levels. Whereas less than 1% of women in the NCI-SEER case-control study were ever exposed to TCE, approximately 7% of women in the Shanghai Women’s Health Study have been exposed, with half of these exposures continuing beyond 1990. A case-control study of lymphoma in Asia would take advantage of the higher prevalence of exposure and the opportunity to link to extensive TCE databases in Asia (e.g., Shanghai CDC database; Guangdong Poison Control Center database). AsiaLymph would also benefit from the use of refined questionnaire workplace modules developed by DCEG investigators to capture chlorinated solvent exposures.

Benzene

Benzene is a ubiquitous occupational and environmental contaminant worldwide, and is used for many applications including pesticides, detergents, and dyes, as well as in the rubber manufacturing process. Although an established leukemogen, there is substantial controversy about its lymphomagenic potential. Both cohort and case-control studies have been somewhat inconsistent (Orsi et al. 2010; Vlaanderen et al. 2010; Cocco et al. 2010, Alexander et al. 2010). There has been some suggestion that this inconsistency may be in part due to the etiological heterogeneity of lymphoma subtypes, as occupational benzene exposure has been shown to be associated with some NHL subtypes but not others (Cocco et al. 2010). A recent meta-analysis of the association between benzene exposure and lymphoid neoplasms found a moderately increased, though not significant, risk of NHL with increasing study quality in workers occupationally exposed to benzene, and acknowledged that the effects of benzene on overall NHL may be attenuated due to the inclusion of specific NHL subtypes that are not associated with benzene exposure (Vlaanderen et al. 2010).

Although DCEG has studied hematopoietic malignancies in a large cohort of Chinese workers with detailed benzene exposure data spanning a 50-year period, the number of NHL cases is limited (i.e., there are ~ 20 benzene-exposed cases with a high probability of being NHL). Because the follow-up period was 1972-1999, almost half of the exposed cases have no specific molecular or histologic information and only nine cases have pathology material for re-review; this lack of biospecimens is the bane of retrospective cohort studies. As such, we have limited power to study the benzene association with NHL and cannot evaluate histology-specific effects. The lack of histologic information and our inability to obtain material from the majority of cases for independent confirmation has raised concerns about our reported benzene-NHL association from the cohort study (Hayes et al. 1997). AsiaLymph will complement the NCI-China CDC benzene cohort study, as the case-control investigation would have substantial power to detect an association between occupational exposure to benzene with confirmed cases of lymphoma (e.g., we would expect ~180 NHL cases with a high probability of exposure to benzene, assuming a prevalence of 3% among controls and an OR of 2.0), would be able to analyze benzene effects on lymphoma by subtype, and would take advantage of the extensive benzene databases we have accumulated on workplace exposures in China.

Environmental exposures to industrial emissions

Environmental pollution has been suspected as a cause of NHL based on findings from occupational studies and increasing incidence in industrialized countries over the past 50 years. Industries of particular interest include petroleum processing (for potential solvent releases) and pulp and paper mills, municipal waste incinerators, and other combustion facilities (for dioxins releases). Studies of NHL in Western countries have found increased risk associated with residential proximity to specific industrial facilities, especially pulp and paper (Linos et al. 1991; Johnson et al. 2003; Ramis et al. 2009), copper smelters (Johnson et al. 2003), and petroleum processing plants (De Roos et al, 2009; Linos et al, 1991). Residence near municipal solid waste incinerators, a major source of dioxin emissions, has been associated with increased risk of NHL in several European studies (Porta et al, 2009). In the NCI-SEER NHL study, we linked residential histories to a nationwide database of dioxin-emitting facilities and observed increased risk associated with residence near cement kilns and hazardous waste incinerators (Pronk et al, submitted). China has undergone rapid industrialization over the past 30 years with little control of industrial emissions until recently (Zhang et al. 2010). In AsiaLymph, we are collecting detailed residential addresses, which will allow us to geocode and map residences over most of participants’ lifetime and to link this information to data on incinerators to estimate environmental dioxin exposure. Most of the study population resides in highly industrialized areas, thus providing an excellent opportunity to follow up on these suggestive findings from Western studies.

Genetic Susceptibility

Lymphomas show significant familial aggregation in the population indicating that genes are likely to play a role in susceptibility. Candidate gene studies have consistently identified SNPs in the pro-inflammatory cytokine, TNF, to be associated with NHL, particularly with DLBCL (Rothman et al. 2006; Fernberg et al. 2010). Genes in other pathways such as DNA repair, oxidative stress, and innate immunity have also been shown to be associated with NHL (Shen et al. 2010; Wang et al. 2006; Hosgood et al. 2011). A major limitation of the TNF results, however, has been the inability to distinguish the TNF association from neighboring human leukocyte antigen (HLA) alleles, which are in linkage disequilibrium (LD) with TNF. Caucasian populations carry the 8.1 ancestral haplotype (AH) that includes the TNF -308A allele (HLA-A1-B8-TNF-308A-DR3-DQ2) (Candore et al. 2002); virtually all individuals with HLA-A*01-B*08-DR*03 have a variant TNF allele (GA or AA). Interestingly, the 8.1 AH is implicated in the risk of numerous autoimmune conditions, including those associated with NHL (e.g., systemic lupus erythematosus, Sjogren’s syndrome) (Candore et al. 2002; Jacob et al. 1990; Newton et al. 2004) and is also associated with higher TNF activity and increased production of autoantibodies. It therefore remains unknown whether the association reported for TNF G-308A is due to or independent from HLA alleles and/or haplotypes.

A large-scale evaluation of genes associated with lymphoma in Asian populations that parallels efforts being conducted in Caucasian populations (e.g., currently a genome-wide association study (GWAS)) would be particularly informative because of Asian population genetic differences in patterns of LD and local haplotype structure (Lan et al. 2007). For example, definitive delineation between TNF G-308A and DLBCL could be achieved in Asian populations where the 8.1 ancestral haplotype does not exist. Among Asians, further delineation from other ancestral haplotypes, notably, 58.1, which includes HLA-A33, -B58, TNF-308A, and -DR3, could also be evaluated (Price et al. 2003). Another example follows from the report by Lan et al. of striking differences between Chinese and Caucasian populations in genotype and haplotype frequencies of polymorphisms in IL4 and IL10 (Lan et al. 2007), which have been associated with NHL among Caucasians (Rothman et al. 2006). To demonstrate the same associations in Asians would add substantial evidence to the causality of these specific SNPs in lymphoma etiology. The study of Asian populations also would allow for identification of novel susceptibility genes.

Viral exposures

Several aspects of lymphoma epidemiology in Asia support the importance of studying potential infectious etiologies for these tumors including the higher incidence of certain types of T-cell lymphomas than in Western countries that are known or likely to be virally-related and the higher prevalence of exposure to certain viruses such as Hepatitis B (Aoki et al. 2008; Aozasa et al. 2008; Du et al. 2009; Kadin et al. 1983). For example, the profound excess incidence of nasal NK/T-cell lymphoma, a uniformly EBV-positive tumor, suggests the existence of important co-factors related to host control of EBV that may be unique to Asian populations (Kadin et al. 1983). Some cases of other histologic types are also EBV-positive, which may be more frequent in T-cell than B-cell derived tumors. An intriguing parallel is nasopharyngeal carcinoma, another uniformly EBV-positive tumor occurring in excess among Asians. Both NK/T-cell lymphoma and nasopharyngeal carcinoma exhibit EBV latency pattern II, characterized by expression of EBV-encoded RNA (EBER), latent membrane proteins, and EBV nuclear antigen-1, but not other EBNAs. We will screen all collected tumor samples by EBER in situ hybridization to identify the EBV-positive lymphomas. Risk factors including demographic, environmental, and genetic characteristics for NK/T-cell lymphoma and for other EBV-positive lymphomas will be compared and contrasted to the risk factors for EBV-negative lymphomas in case-case and case-control comparisons. This will be the most extensive study of EBV positivity and lymphoma carried out to date in an epidemiologic study.

NK/T-cell lymphoma represents a putative EBV-related disorder (Aozasa et al. 2008). We will be collecting an unprecedented number of these tumors for molecular pathologic analysis. In collaboration with our basic science colleagues, we will use microarray-based global gene expression analysis and whole transcriptome deep cDNA sequencing to investigate EBV-specific as well as host genome pathways for this tumor. These studies have the potential to identify proto-oncogene-activating mutations, altered expression of known microRNAs, and/or heretofore uncharacterized NK lymphoma-specific small non-coding RNAs that are dysregulated in NK lymphomagenesis. Results from these collective studies would lead to the identification of candidate lymphoma genes causally involved in NK lymphoma tumor initiation and/or progression.  

As noted, Asians as compared to Western populations have a higher incidence of certain types of T-cell lymphoma (Aoki et al. 2008). The difference is due in part to endemic HTLV-I in southern Japan, but suggests possible involvement in other regions by a second directly transforming agent. This study will generate a large collection of tumors with uniform processing and histologic interpretation. We will select one or more histologic subtypes that appear particularly in excess relative to Western populations. These tumors will be analyzed for evidence of oncogenic viral infections that could explain the excess incidence. We will use high-throughput sequencing of whole genome and whole transcriptome tumor samples followed by digital subtraction analysis to search for non-human sequences other than EBV, including known infections as well as potential novel agents.

Finally, chronic hepatic inflammation caused by hepatitis B (HBV) or hepatitis C viral (HCV) infection has been implicated as a potential risk factor for NHL. The evidence for HCV infection is somewhat more suggestive, although the associated histologic sub-types have not been consistent between studies (Dal and Franceschi 2006). Evidence for HBV is more mixed, with both null (Anderson et al. 2008) and positive associations (Chen et al. 2008; Engels et al. 2010). A NHL study in East Asia, with its relatively high prevalence of HBV infection in particular (Du et al. 2009), provides a valuable opportunity to examine potential important etiologic associations. Accordingly, cases and controls will be screened for Hepatitis B and C exposure and chronic infection. With centralized pathologic review with extensive immunophenotyping, we will have greater precision for examining the associated subtypes and the magnitude of association for each infection. Risk factor analyses for viral-positive cases as well as for histologies with high attributable risk will provide important additional insight into the role of these viruses in lymphomagenesis.

Early life exposures

Emerging evidence suggests that childhood and adolescent exposures in conjunction with genetic makeup may be important in the etiology of NHL. Environmental exposures early in life are important triggers in the development of the adult immune system. The relationship between the development of atopic conditions and surrogates of early-life exposures to infection such as sibship size, childhood crowding, and daycare attendance is well established (Strachan, 1989; 2000). The explanation, termed the ‘hygiene hypothesis’, suggests that delayed exposure to infection leads to subsequent development of atopic conditions via a persistent Th2-dominant immune response or another immune mechanism (Willis-Karp et al, 2001). A recent pooled analysis of 13 case-control studies by the Interlymph consortium (Vadjic et al. 2009) showed significant reductions in B-cell NHL risk among those having at least one atopic condition over their lifetime. The changing social and economic conditions in China, which have resulted in reduced family size and migration from the rural countryside to large urban areas, are likely to provide a broad range in early life exposures to infections. In addition to collecting information about allergies, asthma, and other atopic conditions, we will assess childhood crowding, family size, and early life contact with animals, surrogates of early life infection that have not been extensively evaluated in Asian populations. To date, no large case-control studies have evaluated this hypothesis by histologic type of NHL.

UV Radiation Exposure

Epidemiologic findings generally suggest that exposure to solar ultraviolet radiation (UV) may be associated with a reduced risk of NHL (Armstrong et al. 2007). Increasing ambient UV levels (or, as a proxy, decreasing latitude) have been associated with decreasing NHL incidence or mortality rates in the United States and some parts of Europe (Hartge et al. 1996; Freedman et al. 1997; Grant 2003; Hu et al. 2004), although conflicting ecologic findings have also been reported (McMichael et al. 1996; Bentham 1996; Langford et al. 1998). Several case-control studies from Australia, Europe, and the U.S. have also observed decreasing risks of NHL with increasing self-reported lifetime sun exposure (Hughes et al. 2004; Smedby et al. 2005; Hartge et al. 2006; Weihkopf et al. 2007; Petridou et al. 2007; Soni et al. 2007).Thus far, the existing epidemiologic evidence regarding sun exposure and NHL involves studies conducted in Western, predominantly Caucasian, populations. Replication of these findings in other populations, with potentially different lifestyle correlates of time spent outdoors, would strengthen the inference that the sun exposure-NHL association is real and not attributable to confounding. An inverse association between sun exposure and NHL was observed in a recent small case-control study from Singapore, but more evidence from Asian and other populations are needed (Wong et al. 2010). In Asialymph, we will perform a detailed assessment of past UV exposure, incorporating both self-reported estimates of usual time spent outdoors at different periods of life and satellite-derived estimates of intensity of UV irradiance linked to subjects’ lifetime places of residence using the Total Ozone Mapping Spectrometer (TOMS) database (http://toms.gsfc.nasa.gov). Asialymph will be particularly well suited for investigating sun exposure effects given the wide variability in intensity of UV irradiance expected across study centers, owing to the broad range of latitudes [from 39º N (Tianjin, China) to 22 º N (Hong Kong)].

Other Potential Risk Factors

In addition to the postulated risk factors described above, the AsiaLymph study will enable the evaluation of other suspected and/or novel hypotheses that may contribute to lymphoma risk, including diet, alcohol and smoking habits, sleep quality and duration, dental health, and alternative medicine practices. Specifically, studies conducted in Caucasian populations have suggested that a high intake of certain fruits and vegetables, and possibly fish, may lower the risk of NHL, while high consumption of red meat and some dairy products may increase risk, though the relationships are inconclusive (Skibola et al. 2007). Similarly, studies of the association between smoking and NHL have been equivocal, with some suggesting that smoking may increase the risk of follicular lymphoma specifically or that certain types of tobacco may modify risk (Morton et al. 2005; Stagnaro et al. 2004). Given the increasing prevalence of this exposure in parts of Asia, we will also have an opportunity to evaluate patterns of smoking habits which have been understudied in the Asian population. Consumption of alcohol, particularly red wine, may decrease the risk of NHL, though further exploration is needed (Morton et al. 2005; Briggs et al. 2002). Detailed evaluation of these risk factors will clarify this relationship and extend findings to the Asian population. Given the strong relationship between lymphoma and immune status, we have postulated that sleep quality and duration and prior practice of acupuncture, which is thought to have some immunostimulatory properties, may influence lymphoma risk. Circadian rhythm disruption, part of which is influenced by sleeping patterns, has emerged as a potential risk factor for several cancers, including lymphoma, where night-shift workers may have an increased risk of NHL (Lahti et al. 2008; Davis et al. 2006). Finally, a recent report has suggested that dental health may be associated with NHL risk (Michaud et al. 2008). Our large sample size will enable us to adequately explore this and other novel hypotheses.

Extension of AsiaLymph to enroll cases with myeloid neoplasms

Recent improvements in our understanding of the molecular basis of myeloid leukemias have provided an unprecedented opportunity to provide new insights into the etiology of these conditions, and to compare and contrast risk factors across molecularly defined subtypes of myeloid and lymphoid leukemias and other lymphoma subtypes that have been enrolled using the same methods in the same hospitals and that share the same controls.

Identification of Study Centers

Our initial goal was to have enough study centers and public hospitals to be able to enroll 3,300 lymphoma cases and 3,300 controls in three years; to have a number of centers with a high prevalence of exposure to occupational compounds of interest; to have adequate variation in exposure patterns for particular environmental exposures; and, to the extent possible, to carry out the study in centers and regions where NCI personnel have successfully carried out research previously to be able to take advantage of existing infrastructure and experience.

Centers were considered for inclusion in AisaLymph based initially on NCI study personnel’s familiarity with a particular site, additional information provided through lymphoma and hematological pathologists and clinicians we had come to know in the course of our research in Asia, and a literature search to identify investigators who had carried out descriptive or analytic studies of NHL previously.

Additional issues taken into account were as follows:

1. Availability of local industrial hygienists and occupational health personnel to work with us on the exposure assessment effort;

2. Availability of local epidemiologists in each center;

3. Availability of high quality lymphoma pathologists in a given hospital;

4. Willingness to collaborate with other hospitals in a given center;

5. Willingness to collaborate with NCI on a large, multi-centered effort that required shipment of blood samples to NCI, and shipment of tumor samples to Hong Kong for central pathology review and to NCI for molecular analyses.

Objectives

The primary scientific objectives of the study are to evaluate the etiology of lymphoid and myeloid neoplasms in Asia. The main focus of the study is on chemical exposures, viral exposures, and genetic susceptibility, with central pathology review to characterize effects by histologic subtype. The study will be the largest molecular epidemiology study of lymphoid and myeloid neoplasms ever carried out anywhere in the world, and will offer substantial scientific contributions to the literature.

Specific goals are as follows:

1. Investigate the role of environmental exposure to organochlorines and occupational exposure to trichloroethylene, benzene, and other chemical solvents as well as to other potential occupational exposures;

2. Investigate the role of family history, high-prior candidate genetic variants (e.g., TNF/LTA locus) and emerging findings from genome-wide association studies of lymphoid and myeloid neoplasms in Caucasians, and use state-of-the-art genomics to study genetic variants that may be unique to risk of lymphoid and myeloid neoplasms in Asia;

3. Investigate the etiologic role of EBV, Hepatitis B and Hepatitis C; evaluate potential novel viral agents in T-cell lymphoma; carry out studies to understand pathogenetic mechanisms of NK/T-cell lymphoma;

4. Study other potential determinants of lymphoid and myeloid neoplasms including medical conditions, UV exposure, and other lifestyle factors;

5. Determine the influence of risk factors for lymphoid and myeloid neoplasms overall and by histologic subtype determined by central pathology review;

6. Consider survival studies in hospitals and centers that have follow-up of cases.

Study Design and Methods

Study Design

A hospital-based case-control study design will be used for AsiaLymph. A total of 4,200 incident cases of lymphoma and 4,200 hospital-based controls will be enrolled in Hong Kong, Mainland China, and Taiwan, and an additional 2,000 cases of myeloid leukemias will be enrolled in Hong Kong and Mainland China. The sample size has been revised to take into account loss of cases who are not ultimately found to have a lymphoid or myeloid malignancy or who are not classifiable into a specific subtype after final pathology review, and cases and controls who do not agree to provide a blood sample or from whom a full blood sample cannot be obtained.

All subjects will be directly interviewed. To maximize DNA resources, a buccal cell and blood sample will be collected from all subjects. Case and control identification and interview methods were successfully developed in a hospital-based case-control study in Xuanwei, China, and have been modified and successfully piloted in the AsiaLymph study centers.

There will be a hospital study manager at each hospital, one center study manager in Hong Kong, Taiwan, Chengdu, and Tianjin, and an overall study coordinating center in Hong Kong (Figure 1). A pathology center has also been established in Hong Kong. Study logs and questionnaires will be designed for web-based data transmission and biologic samples will be shipped to the NCI every three months. Receipt will be logged and tracked using the Biological and Environmental Sample Tracking (BEST) System. Study progress reports will be generated weekly using the Study Management System (SMS) and will be transmitted weekly to the NCI.

Case Selection and Enrollment

Eligible cases will be patients at a participating hospital (Figure 1) who are between 18 and 79 years of age at time of initial diagnosis and admitted or treated with incident diagnoses of any lymphoid or myeloid neoplasm including all NHL and Hodgkin disease. Although it is important to understand the etiology of lymphoma in children as well, this undertaking would require additional hospitals, instruments, expertise, and funding that are not currently available to our research team. Adults over the age of 80 are generally among the sickest patients in the hospital and often have multiple comorbidities, which may preclude their participation in an interview of this length. Cases will be permanent residents of the general geographic region that is served by the hospital at the time of diagnosis. Cases will include chronic lymphocytic leukemia/small lymphocytic lymphoma, Waldenström macroglobulinemia, plasmacytoma, multiple myeloma, aggressive NK cell leukemia, cutaneous lymphomas, myeloid neoplasms, and immunosuppression-associated cases (such as HIV, post transplant, Methotrexate use). Cases with previous diagnosis of lymphoma, such as acute lymphoblastic lymphoma, multiple myeloma, chronic lymphocytic leukemia, and non-Hodgkin lymphoma, are ineligible.

An incident case will be defined as a case enrolled into the study within 12 weeks after the date of diagnosis of a lymphoid or myeloid neoplasm. Ideally, a case will be enrolled, interviewed, and provide a blood and buccal cell sample at the time of diagnosis and before receiving any type of therapy.

A rapid case ascertainment system will be established for case identification in participating hospitals, as has been done in previous studies in China conducted by DCEG investigators (see Quality Control section). In each hospital, new cases will be identified within 24 hours of diagnosis. Special arrangements will be made with the management and staff physicians of the hospitals (particularly in the oncology and radiology departments) so that all newly diagnosed cases are quickly reported to the study staff. In addition, daily admission logs will be reviewed to identify probable/suspected cases. The interviewer will approach the patient within 24 hours of the patient being diagnosed with lymphoma or a myeloid neoplasm to determine whether the patient is eligible using a short screening questionnaire. The information includes date of admission, age at date of diagnosis/admission, sex, and current area of residence. This eligibility screening will be collected in a private setting (e.g., family conference room). If the patient agrees to participate, they will be asked to sign the case-control study informed consent form, which includes willingness to be interviewed, to provide access to medical records, to provide a buccal cell and blood sample, and to allow a portion of previously collected pathology material to be made available for additional laboratory studies. The minimum requirement for enrollment in the study is providing consent to participate in the interview or to provide either a blood or buccal cell sample. Subjects who consent to one or more of these items will be enrolled. The interviewer will then carry out the interview in private. If anyone other than the subject is present in the room at any point during the interview (e.g., family member checking in on the subject), the interviewer will have an opportunity to document this in the CAPI near the conclusion of the interview.

If a cancer patient is deemed ineligible, the screening questionnaire with personal identifiers will be discarded, but the remaining portion will be retained. Given that this is a hospital-based case-control study and all effort must be made to enroll subjects while they are still in the hospital, it will be necessary to approach potential subjects within the first 24 hours of diagnosis. While many subjects will not be aware of their diagnosis at this point, study staff at each hospital will be trained specifically on how to approach patients that have just been diagnosed with a serious illness, using approaches and experiences from past studies that the study investigators have been involved in. Study staff will work in close coordination with the treating physicians in order to assess the emotional state of the potential subject and to identify an opportune time to conduct the eligibility screening. Further, conducting the eligibility screening in a private setting will allow the interviewer to better establish rapport with subjects and answer any questions or concerns that may arise.

If a case is missed from enrollment during the initial visit to the hospital, then an attempt should be made to bring the patient back to the hospital within 12 weeks of diagnosis, regardless of whether or not any therapy is scheduled. The case will be approached at the next scheduled follow-up visit to the hospital or contacted by the hospital study manager to arrange a hospital visit if one is not scheduled using the provided contact information.

Additional characteristics of cases will be collected during the interview to ensure that controls can be selected that are comparable to the cases according to key characteristics, including age, sex, and distance they live from their home to the hospital.

A case referred from a non-study hospital/clinic to a study hospital will be eligible for inclusion into the study, if they come from the general geographic region served by the study hospital, and if they are enrolled at the study hospital within 12 weeks of diagnosis. In addition, at a minimum the diagnostic slide and preferably additional unstained slides (see Pathology Review), as well as relevant parts of the hospital record and pathology report plus other diagnostic tests, need to be made available.

Scanned medical records will be used to obtain information related to the confirmation of the diagnosis, hospital admissions and discharges in order to validate previous diagnoses, and other conditions that may influence risk of lymphoma, such as personal and family medical history, history of allergy or autoimmune disease, and information related to viral exposures. In addition, treatment information and health status will be obtained from the scanned medical records in the future for cases who continue to be cared for by physicians at study hospitals. This information will be used in future studies of the determinants of survival.

Control Selection and Enrollment

Controls will be enrolled from patients seen at the participating hospitals. Controls will be individually matched to lymphoid neoplasm cases by hospital, age at date of diagnosis/admission (+/-5 years), sex, and date of admission (within 3 months). Further, all cases and controls must live in the same general geographic region served by the hospital.

Controls will be drawn from patients seen at the same hospital for diseases/conditions that are unlikely to be associated with risk factors under study, such as injuries and selected diseases of the circulatory, digestive, genitourinary, and central nervous system (see Appendix A). Each potential control disease has the same general referral pattern as lymphoma cases to avoid bias. Patients with a history of any lymphoma, including acute lymphoblastic lymphoma, multiple myeloma, chronic lymphocytic leukemia, Hodgkin lymphoma, and non-Hodgkin lymphoma, will not be eligible to serve as controls.

The interviewer will approach the potential control to determine if the patient is eligible using a short screening questionnaire, similar to the form and process used for cases. The information includes date of admission, age at date of diagnosis/admission, sex, and current area of residence, and will be collected in a private setting (e.g., family conference room). If the patient agrees to participate, they will be asked to sign the case-control study informed consent form, which includes willingness to be interviewed, to provide access to medical records, and to provide a buccal cell and blood sample. As with cases, controls who provide consent to participate in the interview or to provide either a blood or buccal cell sample will be enrolled. The interviewer will then carry out the interview in a private setting. If a patient is deemed ineligible, the screening questionnaire with personal identifiers will be discarded, but the remaining portion will be retained.

No more than ~15% of controls enrolled in any hospital can have one type of control disease. Hospital study staff will randomly rotate control selection from different departments that are responsible for various control diseases and will ensure that all such departments fully cooperate with the study.

Participation Rates

Based on the relatively high participation rates we have had in various types of studies in Asia, we expect to have relatively high participation rates (e.g., ~70%). We will spend a substantial amount of time during training on approaches to enhance patient participation and refusal conversion. Further, using the on-line study management system, we will closely follow case capture rates, participation rates, and pre-treatment phlebotomy rates and identify hospitals and interviewers that are low outliers and re-train as needed.

Interview

Both cases and controls will be interviewed within 24 hours after they are identified. A computer assisted personal interview (CAPI) in the local language will ascertain occupational, family, medical, and residential histories. Specific exposure modules in the CAPI will be triggered in response to certain combinations of industry and job title for certain type of occupation solvents (e.g., benzene, OCs, TCE). For each residence, we will ascertain the primary source of drinking water at the home and obtain an accurate address for geocoding to allow for future assessment of environmental exposures. Additional questions will focus on other lymphoma hypotheses that have been developed and discussed bi-weekly by DCEG investigators for the past six months, and will include: history of autoimmune conditions and allergies, height and weight, reproductive and breastfeeding history (to model lifetime organochlorine body burden for women), hair dye use, sunlight exposure, sleep duration/quality, diet and alcohol intake, and childhood crowding, contact with animals, as well as variables used to adjust for socioeconomic status (e.g., education level, household income). We will also collect information on whether or not the admission to the study hospital represents the first physician visit for cases, and if not when and where the case was initially evaluated. It is estimated that the interview will take approximately 90 minutes. In addition, interviews will be recorded and reviewed by study QC personnel to evaluate quality of the interview.

The CAPI is being developed by key NCI investigators in collaboration with Westat and the coordinating center in Hong Kong. The CAPI will trigger occupational exposure assessment modules in the OccIdeas system. OccIdeas is a web-based system designed specifically to assess occupational exposures using more detailed, exposure-oriented job and industry questionnaires (‘modules’, www.occideas.org). A standalone version of the OccIdeas system will be developed for AsiaLymph and programmed onto each of the interviewer tablet PCs. This version of OccIdeas will implement modules developed by NCI that have been refined to ask questions about work tasks with the potential for chlorinated solvent and benzene exposure. A key feature of the OccIdeas system is the ability to program exposure decision rules based on the patterns of responses to one or more questions, which can provide automated exposure assessments. These decision rules will be developed by the NCI and programmed into the standalone version of OccIdeas.

Biological Sample Collection

We plan to obtain a blood sample if at all possible when blood is being collected from subjects for clinical care, so that an additional phlebotomy is not needed. The specific time that blood is usually collected for clinical purposes may vary across the study hospitals, and therefore the study staff will work closely with the clinical care team at each facility to identify consented subjects and coordinate blood collection whenever possible. For example, in hospitals that typically collect blood for clinical care in the early morning, study staff would inform the clinical nurses of consented subjects from the previous day who have not yet provided biologic samples so that extra blood is drawn for research purposes at the time of collection. A 30 ml blood sample will be collected from each study subject. To the extent possible, blood samples will be collected prior to initiation of therapy. The samples will be collected in EDTA vacutainers (3 tubes with 10 ml each). The samples will need to be transported to the processing laboratory within 4 hours and processed, which will include standard low speed centrifugation, vortexing, and aliquotting into 15-1ml plasma aliquots and the remaining blood fraction. Aliquots will then be stored at either -20ºC short-term for up to one week and then stored at -80ºC, or stored at -80ºC immediately after aliquotting. Information about biologic sample collection, processing, aliquotting and storage will be entered directly into a web-based information system as soon as samples are processed. For NK/T lymphoma and CLL cases, based on initial diagnosis, 1 of the EDTA tubes will have mononuclear cells separated and cryopreserved at 2-3 of the study hospitals that have experience with this procedure using controlled rate freezers before being transferred to a liquid nitrogen freezer. Blood samples will be cryopreserved within 6-8 hours of collection. Standardized procedures and training will be provided to each processing lab. The processed samples will be shipped to the study center and subsequently the NCI biorepository, where they will be stored at -80 ºC. Some hospitals may wish to retain up to one-third of blood samples to carry out research. In these instances, the remaining two-thirds of each type of blood sample will be shipped to the NCI biorepository.

A buccal cell sample will be collected and used as an additional source of DNA. Buccal cells will be collected from cases and controls by swishing water in mouth for about a minute. Isopropanol will be added to the sample, which will then be centrifuged, the supernatant removed, and cells frozen. The pellet will be stored at -20 ºC, and samples will subsequently be shipped to the study center followed by the NCI Repository and stored at -80ºC.

Tumor tissues, including pathology blocks for review, diagnosis, and molecular typing, and snap frozen tissues will be collected where possible. For each case, 25 unstained 5 micron sections on HistoBond slides will be made and shipped to the study pathology center. Since the various pathology laboratories do not favor sending out paraffin blocks (due to the duty to keep safe custody of the blocks), unstained paraffin sections will be sent for pathology review instead (10% buffered formalin, if possible). Each participating center will be supplied with HistoBond glass slides. The 25 unstained sections will be cut and mounted on the HistoBond slides (slides are labeled with the original pathology number, with the unique project number beneath it, and a label with the specimen identifier and barcode for tracking in BEST). Two 20 μm-thick sections will also be cut and placed in two separate eppendorfs for molecular studies. These will be sent to the hospital coordinator and delivered to central storage. Tissue slides will be stored at 4 ºC. In addition, a copy of the pathology report plus a copy of all relevant diagnostic tests should be made (e.g., flow cytometry, molecular studies or cytogenetic studies) and transmitted.

The feasibility of cryopreserving lymphocytes in liquid nitrogen in cases and controls, and snap freezing tumor samples from cases, will be assessed in two centers where liquid nitrogen is routinely used.

Study Subject Compensation

Study participants will be compensated about $22.50for completing the interview, donating a buccal sample, and a blood sample. It is estimated that approximately 2 hours will be required for the interview and the collection of buccal cell and blood samples.

Study Organization

The U.S. National Cancer Institute is funding this collaborative, hospital-based case-control study of lymphoid and myeloid neoplasms in four centers in Eastern Asia: Hong Kong, Chengdu, Tianjin, and Taiwan (Figure 1). The NCI study PIs are Drs. Qing Lan and Nathaniel Rothman. Besides DCEG funding, additional support for the study has been obtained at the Hospital and Center level, particularly with regard to making personnel available to carry out the various study components. The study will enroll 6,200 cases and 4,200 controls from approximately mid- 2012 to mid-2018.

The AsiaLymph Coordinating Center Co-PIs, located at the University of Hong Kong, are Drs. T.H. Lam and Dennis Ip. Dr. Lam is a leading epidemiologist in this region with extensive experience conducting international studies in Asia and Dr. Ip has experience conducting epidemiologic studies in Hong Kong. In addition, Dr. John K.C. Chan will be leading the pathology component of the study. He is one of the leading lymphoma pathologists in the world who has contributed to the World Health Organization classification scheme of lymphoma histology. Leaders for the Hong Kong, Chengdu, Taiwan, and Tianjin centers are Drs. Y.L. Kwong, Caigang Xu, Brian Chiu, and Kexin Chen, respectively, who have experience conducting multi-hospital studies. Dr. Roel Vermeulen is the study Co-PI for occupational and environmental exposure assessment and is one of the leading industrial hygienists in the field, with extensive research experience in Asia through work done in collaboration with NCI investigators over the last 10 years.

Figure 1: Overview of the AsiaLymph study organization.

Study Coordinating Center: The overall study coordination in Asia will be by the School of Public Health, University of Hong Kong (Dr. T.H. Lam). Dr. Jun Xu will assist Dr. Lam with the field management of all aspects of the study. The Coordinating Center will manage the activities of the four study centers (Hong Kong, Chengdu, Tianjin, and Taiwan), all participating hospitals, and will coordinate the activities of the study pathology center. Investigators will maintain regular contact with each center and hospital, carry out training and re-training as necessary, receive enrollment data, questionnaires, biologic sample specimen information, and hospital records. They will make regular visits to each center and participating hospital to carry out site-visits and monitor quality control and coordinate shipments of biologic samples.

In addition to coordinating all aspects of the study, the coordinating center staff will be responsible for implementing a computer assisted personal interview (CAPI), which will ascertain occupational, family, and residential histories. The coordinating center will also be responsible for purchasing tablet PCs, and uploading the SMS and CAPI, so they can be used for interviews conducted by the four study centers in Hong Kong, Taiwan, Tianjin, and Chengdu. Questions and question structure used during the CAPI will be provided by the NCI in English. The coordinating center will be responsible for system maintenance and all of the computerized study management components of AsiaLymph.

Study Pathology Center: The pathology component of the study will be coordinated by the Pathology Department, Queen Elizabeth Hospital, Hong Kong (Dr. John K.C. Chan). Investigators there will receive all pathology samples, pathology reports, and relevant medical records and tests and organize and carry out central pathology review (see Pathology Review section). They will also maintain a sample biorepository that will be used to carry out additional analyses.

Center Organization: The study centers will consist of one site in each region that will coordinate and support all aspects of the AsiaLymph study in the hospitals in each respective region. These centers include Queen Mary Hospital (Hong Kong), the West China Hospital of Sichuan University (Chengdu), University of Chicago (Taiwan), and Tianjin Medical University (Tianjin) (Figure 1).

There will be one center study manager in each study center (Hong Kong; Chengdu; Taiwan; Tianjin) who will be responsible for all parts of the study taking place in their center including distribution of supplies and programmed laptops, funding reimbursement to hospitals and physicians for study subject enrollment, coordination of biologic sample storage and shipment to the processing laboratory (Pathology Review Center and/or NCI), coordination of interview data and biological sample collection, maintaining quality control oversight, and other related tasks at each hospital. Further, the center study manager will hire and train a hospital study manager at each hospital in his/her region, and will train and work with the admissions staff at each hospital to identify potential cases and controls. In addition, the center study manager will also assist the hospital study manager in the two largest study hospitals (one in Chengdu and one in Tianjin). Further, the center study manager will train and work with the phlebotomist/buccal sample lab processor at each hospital to collect study patients’ blood specimens and buccal cell samples, following the biological sample collection protocol. The center study manager will maintain regular contact with each study hospital in their region, and will conduct monthly site-visits at each hospital.

In addition, the center study manager will utilize the SMS and the web-based Biological and Environmental Sample Tracking (BEST) System to review study progress and monitor quality control oversight measures from each hospital. This includes routinely collecting and reviewing enrollment data, CAPI data questionnaires, information on samples, and hospital records to ensure complete case and control identification. The center study manager will meet with NCI Investigators every three months as planned by the Study Coordinating Center, and will send monthly reports to the NCI on recruitment, diagnosis, pathology, staff changes, protocol violations if applicable, and quality control measures. The quality control report should include the number of cases and controls enrolled, participation rates, status of the sample shipments, and the completeness of enrollment data, questionnaires, biologic sample specimen information, and hospital records for enrolled subjects. The study center will coordinate shipments of blood and buccal samples to the NCI every 3 months (overnight shipping with tracking). Similarly, it will coordinate shipments of pathology specimens to the study pathology center every 3 months (overnight shipping with tracking). Specimens will be tracked via the specimen tracking system provided by the NCI.

Hospital Organization: Each hospital will have one person who will serve as the hospital study manager, and will have overall responsibility for all parts of the study taking place in their hospital. This includes rapid subject enrollment and interviewing, determining eligibility and documenting consent, collecting and scanning medical records, preparing and shipping pathology materials, coordinating blood and buccal cell collection, processing, and shipment, documenting all study outcomes and activities in the SMS and BEST systems, reporting to the center study coordinator, and other related tasks. All medical records will be scanned and uploaded into the SMS. The hospital study manager will have responsibility for quality control for all parts of the study in their hospital. Each participating hospital will need to have high-speed internet access, as information on each study subject will be entered in real-time directly into the study website and scanned documents and the questionnaire will need to be uploaded on a regular basis. Each hospital additionally will host monthly site visits by the study center manager and meetings with the study coordinating center staff.

Biological Sample Tracking

BEST will track the status of the cases, controls and samples, including the unique project number. Before shipment of the slides, the details of the shipment will be entered and scanned into the system. The shipment manifest will then be printed out to accompany the slides. Cases will be shipped in batches approximately every 2-3 months to the study pathology center at the Department of Pathology, Queen Elizabeth Hospital, Hong Kong.

For cases with limited material, the best opportunity of getting unstained sections for this project is at the time the diagnostic immunohistochemical stains are ordered (with the expectation that most cases of lymphoma will be enrolled in the study). Depending on the size of the tissue, 5-12 additional unstained sections besides those required for the in-house immunohistochemical staining will be cut. The left-over unstained sections will then be shipped out for central pathology review (the pathology center may request the pathologist to bring along original immunostained slides to the consensus conference for review if the available unstained sections are inadequate for full immunophenotyping). For cases with no more tissue in paraffin blocks, original slides and immunostains will not be shipped out, but will be brought to the consensus conference by the participating pathologist of the city. Since leukemia, Waldenstrom macroglobulinemia, and multiple myeloma cases are usually handled by “wet pathologists” and pathologic materials will be limited, 5-10 unstained sections on coated slides taken from marrow biopsies will be shipped out like other cases. Reports (including flow cytometry, cytogenetics, etc.) will be scanned into the web-based system and uploaded to BEST and original slides/smears will be brought to the consensus conference for review. Since participating pathologists do not have experience with diagnosis of these cases, the cases will instead be reviewed by two hematopathologists in Hong Kong.

Pathology Review

Evidence increasingly supports both commonality and heterogeneity in the etiology of lymphoid neoplasms. It is therefore essential that an epidemiologic study of lymphoid or myeloid neoplasms achieve high quality diagnostic specificity in identifying disease subtypes. Classification of lymphoid and myeloid neoplasms has evolved rapidly in recent decades. In 2001, the World Health Organization (WHO) introduced a new classification that was adopted worldwide and represents the current gold standard for classifying all hematopoietic neoplasms (Jaffe et al. 2004; Swerdlow 2009). The WHO classification distinguishes approximately 45 lymphoid neoplasm subtypes based on morphologic, phenotypic, genotypic, immunologic and clinical features, the relative importance of which depends on the specific subtype. WHO subtypes are ideally assigned by an expert hematopathologist after review of diagnostic material and additional clinical and laboratory test results (The Non-Hodgkin's Lymphoma Classification Project, 1997; Jaffe E 2009;Clarke et al. 2004). Because of variability in the laboratory testing undertaken, additional immunophenotyping is often required in order to achieve high confidence in the diagnosis (Turner et al. 2004).

Based on the importance of accurately identifying lymphoid and myeloid neoplasms and classifying disease subtypes, the proposed study will conduct centralized pathology review for all cases using the gold-standard WHO classification. The pathology review will take place in Hong Kong, led by Dr. John Chan and in collaboration with Dr. Dennis Weisenburger, both internationally-recognized expert hematopathologists. For each case, a minimum of 25 x 5-micron unstained slides and diagnostic slides (e.g., hematoxylin and eosin-stained slides) will be sent from each study center to Dr. Chan, accompanied by copies of pathology reports and laboratory tests. The cases (25 unstained sections and printed shipment report for each case) will be shipped in batches (every 3 months) to the pathology review center at Department of Pathology, Queen Elizabeth Hospital, Hong Kong. Specimens will be packed carefully to avoid breakage of the slides. Specifically, after air drying and briefly baking, slides will be thoroughly dried and cooled and all slides of individual cases will be tightly wrapped in paper towels. Adhesive tape with be affixed to the surface, where the unique project number will be written. These rigid “glass blocks” will be tightly stacked and sealed in a card box containing cushion material, and the box will be shaken to ensure no sound is produced (i.e. no dead spaces). Each center will be supplied with the name of a Courier and the customer number for shipment. For the hospitals in Hong Kong, shipments will be delivered through the internal mail system of the Hospital Authority.

Dr. Chan’s laboratory will review the diagnostic materials and accompanying reports, and conduct additional immunophenotyping as necessary to assign the WHO disease subtype. As evidenced by the successful pilot study, we expect a very high percentage of specimen availability (>95% of cases), and transfer of specimens to Dr. Chan’s supervision is acceptable to all study centers. We plan to form a pathology review group with representation from each study center that would meet two times per year during the duration of the study to review pathologic material from the 3,300 cases that will be enrolled into the study over a three-year period.

A consensus conference will be held at Queen Elizabeth Hospital in Hong Kong, and all cases will be reviewed by a panel of experienced pathologists with a consensus diagnosis (full agreement in classification by ≥3 of 4 experts) being reached for each case. The 2008 WHO

Classification of Tumours of the Haematopoietic and Lymphoid Tissues

categories will be used. Unclassifiable cases will not be forced into existing categories, but will be designated “unclassifiable” with notes on why a WHO subtype cannot be assigned. On completion of the project, these cases will be re-reviewed to determine if new entities can be recognized from this group. At the time of pathology review, molecular subtypes of key entities will be classified according to current methods (e.g., identification of DLBCLs by cell of origin).

It is envisioned that each consensus conference will review about 500 cases, and the first one of six is projected to be in July 2016. Logistically, Dr. Dennis Weisenburger and Dr. B. Nathwani, in addition to a panel of pathologists from participating institutions, will review the sorted slides. Since the usual capacity of a pathologist is 50 cases per day, the cases will be divided into two separate sets (250 cases per set), to be examined by two separate panels of pathologists. Cases lacking consensus will be examined by both panels under multihead microscope at the end of the day. During review of the cases, additional immunostains can be requested if necessary to aid in diagnosis/classification, and these will become available the next day. For cases pending further workup, these will be reviewed at the next round. Additional cases to be reviewed at the consensus conference include original slides (H&E and immunostains) of cases for which unstained sections are not available, and original slides (H&E and immunostains) of cases with limited available unstained sections. Original slides and all related reports for leukemia and myeloma cases will be reviewed by two hematopathologists in Hong Kong. Since each panel should consist of 4 pathologists, a consensus conference theoretically will require the participation of 8 pathologists (2 panels, with 4 pathologists each). To economize on the number of invited pathologists, all cases will be reviewed by Dr. John Chan beforehand, and the diagnosis/classification will count towards one of the diagnoses for consensus purposes. Thus, only 6 invited pathologists (2 panels, 3 pathologists each) will be required each time. After each consensus conference, the consensus diagnosis and immunophenotype will be entered into the Study Management System.

Slides not required for WHO subtype classification will be stored for future molecular studies. Examples of such studies include evaluation of protein expression, chromosomal translocations, viruses such as Epstein-Barr virus, and tumor DNA to investigate somatic alterations. Archiving of tumor tissue specimens will also enable the study to take advantage of the rapidly-improving technology for formalin-fixed, paraffin-embedded tissues (e.g., gene expression and microRNA profiling), thus allowing us to be responsive if it is appropriate to newly developed technology at the time the study has completed recruitment. Initially, 50% of tissue samples from each case will be shipped to NCI and the remaining will be stored at Queen Elizabeth Hospital. Additional samples will be shipped to NCI depending on future assay needs.

In selected study centers with existing, more advanced capacity for handling tumor tissues (e.g., Hong Kong), we will explore the possibility of creating tissue microarrays from tumor blocks in order to minimize future laboratory assay costs, as well as obtaining snap-frozen tissues to enable future studies of gene expression.

Linking Occupational Histories with Occupational Exposure Databases

Occupational histories and data from solvent modules will be reviewed by DCEG and local industrial hygienists in each of the study centers, and linked with TCE and benzene exposure databases in Eastern Asia to obtain estimates of TCE and benzene exposure for each job, industry, and region. For example, the Shanghai CDC occupational exposure database contains measurements of TCE beginning in 1963 and benzene measurements since the 1950s from many types of workplaces and is generalizable to factory conditions in China during this period. Further, the NCI-China CDC benzene cohort study also has an extensive database of benzene measurements from many factories in 12 cities going back to the 1950s. Industrial hygienists in Taiwan are currently assembling data from several sources for use in AsiaLymph.

We will extract exposure data from several national, regional, and local institutions into a single exposure database. Experience of such an effort has been obtained in several DCEG projects, including the Shanghai Women’s Health Study, the China Benzene Cohort Study, and the NCI-SEER NHL study. We will also use multi-level modeling of the exposure data where one can extrapolate information of job related exposures across regions and countries, while still accounting for regional differences. A similar approach has been followed in the ongoing NCI-China CDC benzene cohort study, which included 12 different centers. Exposure modeling of the benzene data indicated that there are regional differences in exposures but that the relative ranking of jobs remains largely the same across the different regions. These and other analyses show that routinely collected exposure data from different regions can be used in a statistical framework which allows extrapolation of data across time and geographic regions based on observed similarities in the exposure data (Dosemeci et al. 1997).

Besides the quantitative data, we will also use targeted job and industry questionnaire modules to focus on a few specific exposures (e.g. benzene, TCE). The subject-specific information obtained in these modules will further aid in the determination of how to extrapolate exposure levels across countries and to allow differences in exposure estimates for jobs within a single region/time period.

Quality control

The study team has had extensive experience conducting multi-center studies in Asia over many years. We have studied lung cancer in four cities in Yunnan Province as part of the hospital-based Xuanwei study of lung cancer over the past 5 years; we have studied leukemia, lymphoma, and lung cancer in 12 centers in the benzene cohort study in Mainland China over the past 24 years where subjects speak Mandarin and/or Cantonese; and most recently we have studied the genetics of lung cancer among nonsmoking Asian females in 19 centers in Asia including Mainland China and Taiwan as part of our GWAS replication study. The study team also has members who worked extensively on the multi-center DCEG Brain Cancer study in the United States and the Spanish Bladder Cancer study, carried out in 5 regions and 18 hospitals in Spain. We have used the experience acquired from these previous investigations to design the quality control component of the study. In addition, investigators in Asia have expressed their commitment to the success of the study, given their interest in clinical outcomes and their ongoing involvement in InterLymph.

We will have experienced personnel on site in Asia who will work with us to coordinate and manage the study. Dr. T.H. Lam and Dr. Dennis Ip will be the lead epidemiologists at the coordinating center in Hong Kong. We have recruited Dr. Jun Xu to manage the study from Hong Kong and to have responsibility in Asia for all aspects of quality control. Dr. Xu successfully managed many components of the field phase of the DCEG multi-center hospital-based case-control study of lung cancer in Yunnan Province from 2005-2010. He was trained by DCEG personnel, and has done an outstanding job overseeing training sessions, reviewing quality control, and making regular site-visits to each of the study hospitals in 4 cities in this region. After a one year stay at NCI from early 2010- early 2011 developing the study protocol and instruments, he is moving to Hong Kong in May, 2011 and will work full-time on AsiaLymph at the University of Hong Kong. He will make regular visits to each center and hospital (once every 3 months) throughout the 3-year course of the study to provide oversight.

There will also be a substantial NCI presence on-site reviewing all aspects of the study. Within 3 months after the study begins, the NCI Co-PIs and Dr. Wei Hu, a staff scientist in DCEG, will visit each center and hospital. Dr. Hu also managed many components of the Xuanwei lung cancer study in China and has received additional training in study management and quality control since joining DCEG in 2010. Dr. Hu will visit each center and hospital every 6 months and the NCI study Co-PIs will visit each center every 6 months and selected hospitals annually. We have used this approach successfully in the hospital-based case-control study of lung cancer in Xuanwei. More frequent and targeted visits will be made by Dr. Xu and NCI personnel as necessary to ensure the highest quality control in the study.

In addition to the oversight of the study in Asia, the NCI study Co-PIs plus additional OEEB staff will review the study status on a weekly basis in conjunction with personnel from Westat, including review of enrollment reports generated from the study management system (described below) in Hong Kong, review of questionnaires for completeness and quality, and review of biological sample collection, processing and storage through the Westat-maintained biological sample tracking system. OEEB investigators will also have a weekly phone call with the coordinating center at the beginning of the study and then as the study progresses bi -weekly calls. There will also be annual meetings of all key study personnel from Asia and NCI in Hong Kong for study review and updates.

Finally, we have been able to utilize the same Westat personnel that developed the study materials for the NCI Spanish Bladder Cancer study (a hospital-based case-control study of bladder cancer in 5 study centers, with a total of 18 hospitals) for AsiaLymph. Specifically, Westat personal, under the direction of NCI investigators and with NCI developed protocols, will develop the Manual of Operating Procedures (MOP) for each center and training materials for study management. The training materials are the necessary directions and checklists of tasks for each study center’s coordinator to carry out during the implementation of the AsiaLymph protocol. Training materials for study management as well as for biological sample collection will be distributed to each study center. A study management system (SMS) with online access will also be developed by the Westat under the direction of the NCI, and will be used to monitor the study progress. Specifically, the SMS will track subject enrollment, interview and hospital record status, biospecimen and tumor tissue collection status, and other key data as described in the protocol. The SMS must be available to study staff at the Study Centers for data input and reporting, and to NCI investigators to monitor study progress. The SMS will generate regular reports to the NCI pertaining to the number of cases and controls enrolled, participation rates, the status of sample shipments, and the completeness of the enrollment data and questionnaires. In addition, hospital records and information on biologic specimens will be uploaded into the SMS. Westat personal will also developing training materials, which are based on those developed by Westat for the Spanish Bladder Cancer Study.

Anticipated Distribution of Enrolled Cases

Distribution of lymphoma types expected: Of the 4,200 lymphoma cases we will enroll into AsiaLymph, we anticipate enrolling at least 3,000 pathologically confirmed NHL cases with a full blood sample. Most of the remaining 300 cases will be multiple myeloma cases. Among the NHL cases, we expect 1,400 DLBCL cases and 600 total T-cell lymphomas (of which approximately 200 will be NK/T-cell). The expected distribution of cases is shown in Figure 2. Of the 2,000 myeloid neoplasm cases we will enroll, we anticipate enrolling at least 1,400 pathologically confirmed acute myeloid leukemia (AML) cases with a full blood sample.

Figure 2. Anticipated NHL case enrollment in AsiaLymph

Biological Sample Analysis

Biomarker measurements: The major classes of biomarkers that will be measured in AsiaLymph include plasma markers of long-term exposure to organochlorines and hepatitis B and C infection and active viral replication, DNA-based markers of genetic susceptibility and tumor markers at the DNA, mRNA, and protein level.

A total of 14 ml of plasma will be analyzed for PCBs, DDT and its metabolites, chlordane metabolites, and at no additional cost other pesticides routinely measured in blood samples, and selected co-planar PCBs, dibenzofurans, and dioxins on a subgroup of study subjects, after pilot testing of samples from controls in each study center. The primary approach we are using to assess environmental exposure to PCBs, DDT, other pesticides, and dibenzofurans is biological monitoring because the primary route of environmental exposures in the general population is through dietary contamination, which cannot be reliably estimated by other methods. We reported statistically significant findings for several key organochlorine compounds in a study of 100 cases and 100 controls in the NCI-SEER NHL study (De Roos et al. 2005). We have conservatively budgeted for analyzing samples from 150 cases and 150 controls in AsiaLymph. We will refine the sample size needs after reviewing the results of the above pilot study and note that it is possible that we may need to modestly expand this component of the study. In addition, the subgroup for organochlorine analysis will be limited to cases with pre-treatment blood samples, to controls who similarly have not undergone treatment that can potentially cause local or systemic toxicity or inflammatory responses, and to both cases and controls that have not undergone recent weight loss for any reason.

Viral Assays: Screening hepatitis B and C assays are routinely carried out at some centers and will be made available to us at no cost for case-control analysis (on approximately 1,200 cases and 1,200 controls). We plan on testing the remaining unscreened samples for antibodies to hepatitis B core antigen and hepatitis C at minimal cost in Asia. Importantly, antibody positivity stably indicates viral exposure, with equal reliability in cases and controls, and is unlikely to be affected by disease. All positive samples will then be analyzed for hepatitis B surface antigen and hepatitis C RNA, respectively, which generally measure chronic replication. These assays also can potentially become positive from reactivation due to waning immunologic competence in some lymphomas. Positive results with the latter assays would be suggestive but would need to be confirmed in the future when prospective cohort specimens become available. On the other hand, null results with hepatitis B surface antigen and hepatitis C RNA would be informative in this well-powered study and unlikely to be affected by disease status. Sortable lymphocytes from cryopreserved blood samples will be used for PCR-based analyses of integrated/episomal viral genomes.

Genetic Studies: DNA will be extracted and analyzed for a series of germ-line association analyses. Initially, we will attempt to replicate the candidate gene and other GWAS findings from Caucasian populations in our sample of Asian cases and controls. Any functional variants identified in previous studies will be a high priority to replicate in this population with a sample size of this magnitude. In addition, we will conduct a genome-wide scan for selected subtypes beginning in FY16.

Tumor Sample Analyses

All tumor samples will be re-stained by hematoxylin and eosin and immunophenotyped at Queen Elizabeth Hospital. Special diagnostic stains (immunohistochemical and in-situ hybridization) will be performed including basic panel plus additional stains depending on the individual case, with a sequential strategy to preserve the maximum number of slides for molecular assays. Remaining slides will be wrapped in aluminum foil, and stored according to the unique project number in -80^oC refrigerator to preserve antigenicity. All tumor samples will be screened for EBV-encoded RNA (EBER) by in situ hybridization to identify the EBV-positive lymphomas. For all cases of Burkitt lymphoma or suspected Burkitt lymphoma, FISH will be conducted to detect MYC break-apart and MYC-IGH. For all other cases, FISH will be performed only when required (i.e. BCL2 break-apart, BCL6 break-apart, CCND1 break-apart).

Special studies will be carried out for subtypes of interest, and may include the following, depending on number of cases and availability of pathology samples and resources:

• NK/T-cell studies: Whole transcriptome deep cDNA sequencing will investigate EBV-specific as well as host genome pathways for NK/T-cell lymphomas.

• T cell tumors: For one or more T-cell tumors that appear to be at excess relative to rates in the West, we will use high-throughput sequencing of whole genome and whole transcriptome tumor samples followed by digital subtraction analysis to search for non-human sequences other than EBV, including known infections as well as potential novel agents.

• CLL/SLL: Additional DNA-based studies will be determined.

• Multiple myeloma cases: Collection of unstained bone marrow biopsies will allow characterization of the tumor microenvironment.

Additional studies (immunohistochemical/ FISH) can be performed on this superb collection of cases as unstained sections will be available, including tissue microarray block production which will be considered in the future.

Data Analysis and Power

Statistical Analysis: All study data will be compiled and cleaned at the University of Hong Kong under the direction of study investigators. Analytic files will be maintained at IMS for convenient analysis by DCEG scientists. Since lymphoma comprises a group of related yet heterogeneous diseases, each characterized by the malignant transformation of lymphoid cells but with distinctive morphologic, immunophenotypic, genetic, and clinical features, we will analyze risks by lymphoma subtype as well as larger subgroups.

For analyses by subtype, odds ratios (ORs) and 95% confidence intervals (CIs) will be derived for each risk factor from polytomous unconditional logistic regression models. P values for the linear trend will be computed for continuous variables and using ordinal variables. To evaluate heterogeneity among lymphoma subtypes, we will use 2 statistical approaches. First, we will conduct a homogeneity test in the polytomous model, testing the null hypothesis that the regression coefficient for each risk factor was the same for all subtypes. Values of P less than .05 will be considered to provide evidence of heterogeneity. The test for homogeneity has the greatest power to detect risk differences when the risks for the subtypes all vary slightly from one another. Second, we will analyze all possible case-case pairwise comparisons using dichotomous logistic regression models (Morton et al. 2008). We will compute test the null hypothesis that the particular risk factor does not discriminate between the 2 disease groups modeled. To account for the pairwise analysis, we will apply a Bonferroni correction. In contrast to the test for homogeneity, the pairwise analysis has the greatest power to detect risk differences when the risk for one disease group is distinct from the other(s). For risk factors with more than 2 categories, we will use the ordinal variable for the homogeneity test and pairwise analysis. We will also explore additional statistical approaches currently under development in InterLymph for investigating etiologic differences among NHL subtypes. Analyses will also be conducted for larger lymphoma subgroups including NHL and B-cell lymphomas, and all myeloid neoplasms and AML, using logistic regression models. For genetic analyses, standard methods will be used to test the effect of each SNP. We will also have the opportunity to conduct exploratory analyses using a new powerful and flexible subset-based approach to the combined analysis of heterogeneous traits, which is an approach that agnostically explores subsets of the traits to identify the strongest association signal and then evaluates the significance of the detected association using efficient adjustment for multiple correlated tests involved (N. Chatterjee, personal communication).

Initial analyses will be conducted for lifestyle risk factors, occupational exposures, environmental exposures, viral exposures, and genetic main effects. Exploratory gene-environment interaction analyses will also be conducted. We will also conduct genetic pathway analysis to evaluate whether the set of genes in a well-defined pathway (e.g., Th1/Th2 pathway) are associated with the disease risk. This type of analysis is particularly helpful in situations when the pathway is enriched with multiple SNPs with small effects. All models will be adjusted for sex, age, study center, and date of enrollment (the control matching factors) and education. Additional potential confounders will be selected based on initial analyses of the study data set and through identification of well-established risk factors in the literature.

Power Analysis: For a dichotomous exposure variable with a prevalence of 2%, we will have 80% power (two-sided alpha = 0.05) to detect ORs of 1.40, 1.51 and 1.72 for all NHL, DLBCL and AML, and total T-cell lymphomas, respectively, using 3,300 controls in a logistic regression model (Table 1). For an exposure variable with a prevalence of 3%, we will have 80% power to detect ORs of 1.32, 1.41, and 1.58 for each of these case groups, respectively. For risk factors with high prevalence of exposure, we will have adequate power to detect lower odds ratios.

For studies of genetic polymorphisms, we will have 80% power to detect an OR of 1.2 per allele (from an additive genetic model) for minor allele frequencies (MAFs) of 8%, 13.5%, and 36% for all NHL, DLBCL and AML, and total T-cell cases, respectively (Table 2). We will be able to detect an OR of 1.3 per allele for MAFs of 3.5%, 5.5% and 11.5% for these categories, respectively.

Table 1. Power table according to exposure prevalence

Exposure Prevalence	NHL Subtype	Odds Ratio	Power
2%	NHL	1.40	80%
	DLBCL	1.51	80%
	Total T-Cell	1.72	80%
3%	NHL	1.32	80%
	DLBCL	1.41	80%
	Total T-Cell	1.58	80%
5%	NHL	1.25	80%
	DLBCL	1.32	80%
	Total T-Cell	1.45	80%
10%	NHL	1.18	80%
	DLBCL	1.23	80%
	Total T-Cell	1.32	80%

Table 2. Power table according to minor allele frequency (MAF)

Odds Ratio per Allele	NHL Subtype	MAF	Power
1.2	NHL	8%	80%
	DLBCL	13.5%	80%
	Total T-Cell	36%	80%
1.3	NHL	3.5%	80%
	DLBCL	5.5%	80%
	Total T-Cell	11.5%	80%
1.4	NHL	2%	80%
	DLBCL	3.5%	80%
	Total T-Cell	6.5%	80%
1.5	NHL	2.5%	80%
	DLBCL	2%	80%
	Total T-Cell	4%	80%

Personnel

Co-Principal Investigators

Qing Lan (OEEB), Nathaniel Rothman (OEEB): Overall study conduct

Lead Investigators

Lindsay Morton (REB): Pathology review and tumor molecular analyses

Martha Linet, Alina Brenner (REB): Study design

Charles Rabkin (IIB): Viral studies

Mark Purdue (OEEB): Occupational solvent exposures

Lynn Goldin, Neil Caporaso (GEB): Family history, CLL studies

Stephen Chanock (CGR, LTG): Genomics

John Chan (Queen Elizabeth Hospital): Pathology

Roel Vermeulen (Utrecht Univ.), Melissa Friesen (OEEB): Occupational exposure assessment

Mary Ward (OEEB): Environmental exposure assessment

Nilanjan Chatterjee, Kai Yu (BB): Study design and statistical analysis

Wei Hu (OEEB), Bryan Bassig (OEEB) and H. Dean Hosgood (Einstein Medical School): Field training, quality control, and study management at NCI

Xu Jun (Univ. of Hong Kong): Field study management in Asia

Additional Intramural Co-Investigators

Sonja Berndt (OEEB), Butian Ji (OEEB), Allan Hildesheim (IIB), Neil Caporaso (GEB), Meredith Yeager (CGR), Eric Engels (IIB)

Co-Principal Investigators in Asia

T.H. Lam and Dennis Ip (University of Hong Kong)

Study Center Leaders

Y.L. Kwong (Hong Kong), Brian Chiu (Taiwan), Caigang Xu (Chengdu), Kexin Chen (Tianjin)

Consultants

Debra Silverman (OEEB), Lin Fritschi (Univ. of Western Australia), Troy Sadkowsky (Univ. of Western Australia), Sophia Wang and Dennis Weisenburger (City of Hope)

Human Subjects Protection

Briefly, a trained hospital coordinator will identify potentially eligible cases and controls and approach the patient to discuss the study using the initial contact script. The coordinator will make it clear to the patient, in a non-coercive environment, that participation in this study is strictly voluntary and their healthcare will in no way be affected regardless of study participation. The study manager will then determine the patient’s eligibility. If the patient agrees to participate, informed consent will be requested after a full explanation of the benefits and hazards of participation in the study. All information received through the eligibility and consenting processes will be kept strictly confidently regardless of study participation. Written consent will be obtained for each of the components of the study: (1) access to medical records and pathology material, (2) interview, (3) buccal cell and blood sample collection. The patient can give consent to some or all of these components. Upon consenting, trained staff will administer the interview and collect the biological samples. All original interview responses and other study documents with personal identifiers will be kept securely at the study center. All interview responses and other study documents will be kept strictly confidential. The National Identifying Number of each subject will be collected in the CAPI as this information is typically collected in the hospital setting. This number along with other personally identifiable information will be kept strictly confidential. Data collected on study computers will be password protected and encrypted, with access limited to qualified and trained members of the study staff. To further protect patient confidentiality, each hospital will have a designated area where study records and equipment will be securely stored (i.e. a locked room or filing cabinet). All data from the CAPI that is uploaded into the SMS will be encrypted and uploaded and stored using secure and password protected servers. All confidential patient identifiers in the CAPI, including the National Identifying Number and address of the subject, will be removed by the coordinating center before the data is sent to the NCI. Information will be kept at the coordinating center for quality control purposes, but no other investigators will have access to data including patient identifiers.

All data from the CAPI will be uploaded to the SMS on a daily basis. An archived copy of the data will be stored on the tablets in a password protected and hidden folder that only the study PIs have access to until confirmation that the uploading of information was successful. Medical records that are scanned and uploaded into the SMS will first be masked by study staff using labels to cover all personally identifiable information in the records. Study staff will work with each hospital to evaluate the format of the medical records and to identify all locations within the records where identifying information might be found. This will assist in the proper training of the person responsible for the masking procedure by study investigators, with oversight from the study manager in each hospital.

All biological samples will contain study ID numbers only, and all samples sent from the NCI biorepository to laboratories for analysis will be characterized by identifier numbers only (i.e. no patient information). Biological sample analytic results will be sent to the NCI for addition to the study database. There will be no personal identifiers in data analysis files prepared by the study coordinating center for study investigators. Finally, no individual results will be presented in publications or other reports. All procedures for this study will be conducted according to the recommendations of the World Medical Association Declaration of Helsinki for human study subject protection. At any time during the study, subjects who wish to discontinue their participation in the study may do so, and further may request to withdraw consent for use of any collected data, including medical records. The name and contact information for the relevant authority at each hospital is provided on each of the consent forms.

A Federal-Wide Assurance will be established with each study center. The protocol will also be reviewed by the NCI SSIRB, the Westat IRB, and IRBs from collaborating hospitals.

Recruitment

Case-control study: Subjects will be recruited by on-site collaborators in each participating hospital. The hospital coordinator will identify eligible cases from daily admission records and approach the patient to discuss the study and then, if the patient agrees to participate, carry out informed consent. There will be consent for access to medical records and pathology material, interview, buccal cell and blood sample collection. In addition, cases will be consented for follow-up of medical treatment and clinical status.

The hospital coordinator will identify eligible controls based on age, and reasons for admission from selected Departments. They will approach each potential control, briefly explain the study, and a short screening questionnaire will be administered to determine eligibility. We will then use this information to select controls for the study, who will be administered the full questionnaire and be asked to provide biological samples as described above for cases. All patients who are eligible will be enrolled as controls.

Informed Consent: Informed consent will be requested, in their local language, after a full explanation of the benefits and hazards of participation.

Potential benefits and risks: No direct benefits to the participants are expected from this study except the satisfaction of contributing to the scientific understanding of etiology of lymphoma. The research involves no more than minimal risk to subjects from venous phlebotomy which will be carried out by local medical or nursing staff. No physical harm is expected from the collection of buccal cells. No other risks are expected from participating in this study. Participation is voluntary.

Compensation: For the case-control study, subjects will be compensated $22.50 for time and effort spent in this study. It is estimated that up to 2 hours will be required from subjects during the course of this study.

Communication of study results: Pathologists will be notified of the final lymphoma classification. In cases where the central pathology review differs from the initial hospital pathology diagnosis, the hospital pathologist who did the initial review will be informed of the consensus diagnosis by the chair of the central pathology review group. We do not plan to provide individual results to study subjects or their physicians because the assays are for research only and have uncertain clinical relevance. The laboratory results will be used to understand the etiology of lymphoma in humans. The research laboratories will use state-of-the art methods that can be duplicated, but most of the protocols will not be used or approved for clinical settings. Individual requests for study data will be honored, as required by law. The risks to study subjects are minimal. The study results could help us to understand the etiology of lymphoma in this population and elsewhere.

Confidentiality of study data: All original questionnaire responses, other study documents and medical record abstraction data with personal identifiers will be kept securely at the study coordinating center. All biological samples sent from the NCI biorepository to laboratories for analysis will be characterized by identifier numbers only. Biological sample analytic results will be sent to for addition to the study database. All personal identifiers will be stripped from data analysis files prepared by the study coordinating center for study investigators. Finally, no individual results will be presented in publications or other reports.

Appendix A: Eligible control diseases

Controls will be drawn from multiple non-malignant disease categories that have not been linked to risk factors under study and are not known or suspected to have an immunological, infectious, and/or inflammatory etiology.

1. Injuries

   1. Fractures (800-829)

   2. Injury to blood vessels (900-904)

2. Diseases of the circulatory system

   1. Hypertensive disease (401-405)

   2. Ischemic heart disease (410-414)

   3. Diseases of pulmonary circulation (415-417)

   4. Cerebrovascular disease (430-438)

3. Diseases of the digestive system

   1. Appendicitis (540-543)

   2. Hernia of abdominal cavity (550-553)

   3. Other diseases of intestines and peritoneum (560-569) – intestinal obstruction, diverticula, etc.

   4. Other diseases of digestive system (574) – cholelitiasis

4. Diseases of the genitourinary system

   1. Other diseases of urinary system (591, 592-594, -599) – calculus of kidney and ureter

5. Diseases of the central nervous system and sense organs

   1. Other disorders of the central nervous system (340-349) – hemiplegia/hemiparesis, epilepsy and current seizures

Reference List

The Non-Hodgkin's Lymphoma Classification Project. 1997. A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. Blood 89:3909-18.

Alexander D, Wagner ME. 2010. Benzene exposure and non-Hodgkin lymphoma: a meta-analysis of epidemiologic studies. JOEM 52: 169-189.

Anderson LA, Pfeiffer R, Warren JL, Landgren O, Gadalla S, Berndt SI, Ricker W, Parsons R, Wheeler W, Engels EA. 2008. Hematopoietic malignancies associated with viral and alcoholic hepatitis. Cancer Epidemiol Biomarkers Prev 17:3069-3075.

Aoki R, Karube K, Sugita Y, Nomura Y, Shimizu K, Kimura Y, Hashikawa K, Suefuji N, Kikuchi M, Ohshima K. 2008. Distribution of malignant lymphoma in Japan: analysis of 2260 cases, 2001-2006. Pathol Int 58:174-182.

Aozasa K, Takakuwa T, Hongyo T, Yang WI. 2008. Nasal NK/T-cell lymphoma: epidemiology and pathogenesis. Int J Hematol 87:110-117.

Armstrong BK, Kricker A. 2007. Sun exposure and non-Hodgkin lymphoma. Cancer Epidemiol Biomarkers Prev 16: 396-400.

Au WY, Lo J. 2005. HTLV-1-related lymphoma in Hong Kong Chinese. American Journal of Hematology 78:80-81.

Bentham G. 1996. Association between incidence of non-Hodgkin's lymphoma and solar ultraviolet radiation in England and Wales. BMJ 312:1128-1131.

Briggs NC, Levine RS, Bobo LD, Haliburton WP, Brann EA, Hennekens CH. Wine drinking and risk of non-Hodgkin’s lymphoma among men in the United States: a population-based case-control study. 2002. Am J Epidemiol 156: 454-462.

Candore G, Lio D, Colonna RG, Caruso C. 2002. Pathogenesis of autoimmune diseases associated with 8.1 ancestral haplotype: effect of multiple gene interactions. Autoimmun Rev 1:29-35.

Centers for Disease Control and Prevention. National Report on Human Exposure to Environmental Chemicals: Organochlorine Pesticides. 2010. http://www.cdc.gov/exposurereport/data_tables/chemical_group_0802.html. Accessed 10/22/10.

Chen MH, Hsiao LT, Chiou TJ, Liu JH, Gau JP, Teng HW, Wang WS, Chao TC, Yen CC, Chen PM. 2008. High prevalence of occult hepatitis B virus infection in patients with B cell non-Hodgkin's lymphoma. Ann Hematol 87:475-480.

Chia KS, Du WB, Sankaranarayanan R, Sankila R, Seow A, Lee HP. 2001. Population-based cancer survival in Singapore, 1968 to 1992: an overview. Int J Cancer 93:142-147.

Choi WW, Weisenburger DD, Greiner TC, Piris MA, Banham AH, Delabie J, Braziel RM, Geng H, Iqbal J, Lenz G, Vose JM, Hans CP, Fu K, Smith LM, Li M, Liu Z, Gascoyne RD, Rosenwald A, Ott G, Rimsza LM, Campo E, Jaffe ES, Jaye DL, Staudt LM, Chan WC. A new immunostain algorithm classifies diffuse large B-cell lymphoma into molecular subtypes with high accuracy. Clin Cancer Res. 2009;15:5494-502.

Clarke CA, Glaser SL, Dorfman RF, Bracci PM, Eberle E, Holly EA. 2004. Expert review of non-Hodgkin's lymphomas in a population-based cancer registry: reliability of diagnosis and subtype classifications. Cancer Epidemiol Biomarkers Prev 13:138-143.

Cocco P, t’Mannetje A, Fadda D, Melis M, Becker N, de Sanjose S, Foretova L, Mareckova J, Staines A, Kleefeld S, Maynadie M, Nieters A, Brennan P, Boffetta P. 2010. Occupational exposure to solvents and risk of lymphoma subtypes: results from the Epilymph case-control study. Occup Environ Med 67: 341-347.

Colt JS, Rothman N, Severson RK, Hartge P, Cerhan JR, Chatterjee N, Cozen W, Morton LM, De Roos AJ, Davis S, Chanock S, Wang SS. 2009. Organochlorine exposure, immune gene variation, and risk of non-Hodgkin lymphoma. Blood 113:1899-1905.

Colt JS, Severson RK, Lubin J, Rothman N, Camann D, Davis S, Cerhan JR, Cozen W, Hartge P. 2005. Organochlorines in carpet dust and non-Hodgkin lymphoma. Epidemiology 16:516-525.

Conde L, Halperin E, Akers NK, Brown KM, Smedby KE, Rothman N, Nieters A, Slager SL, Brooks-Wilson A, Agana L, Riby J, Liu J, Adami HO, Darabi H, Hjalgrim H, Low HQ, Humphreys K, Melbye M, Chang ET, Glimelius B, Cozen W, Davis S, Hartge P, Morton LM, Schenk M, Wang SS, Armstrong B, Kricker A, Milliken S, Purdue MP, Vajdic CM, Boyle P, Lan Q, Zahm SH, Zhang Y, Zheng T, Becker N, Benavente Y, Boffetta P, Brennan P, Butterbach K, Cocco P, Foretova L, Maynadié M, de Sanjosé S, Staines A, Spinelli JJ, Achenbach SJ, Call TG, Camp NJ, Glenn M, Caporaso NE, Cerhan JR, Cunningham JM, Goldin LR, Hanson CA, Kay NE, Lanasa MC, Leis JF, Marti GE, Rabe KG, Rassenti LZ, Spector LG, Strom SS, Vachon CM, Weinberg JB, Holly EA, Chanock S, Smith MT, Bracci PM, Skibola CF. 2010. Genome-wide association study of follicular lymphoma identifies a risk locus at 6p21.32. Nat Genet 42: 661-664.

Dal ML, Franceschi S. 2006. Hepatitis C virus and risk of lymphoma and other lymphoid neoplasms: a meta-analysis of epidemiologic studies. Cancer Epidemiol Biomarkers Prev 15:2078-2085.

Davis S, Mirick DK. Circadian disruption, shift work and the risk of cancer: a summary of the evidence and studies in Seattle. 2006. Cancer Causes Control 17: 539-545.

De Roos AJ, Hartge P, Lubin JH, Colt JS, Davis S, Cerhan JR, Severson RK, Cozen W, Patterson DG, Jr., Needham LL, Rothman N. 2005. Persistent organochlorine chemicals in plasma and risk of non-Hodgkin's lymphoma. Cancer Res 65:11214-11226.

De Roos AJ, Colt JS, Blair A, Airola M, Severson R, Cozen W, Cerhan JR, Hartge P, Nuckols JR, Ward MH. Residential proximity to industrial facilities and risk of non-Hodgkin lymphoma. Environ Research 110:70-78; 2010

Dosemeci M, Rothman N, Yin SN, Li GL, Linet M, Wacholder S, Chow WH, Hayes RB. Validation of benzene exposure assessment. Ann N Y Acad Sci. 1997 Dec 26;837:114-21.

Du F, Liu QL, Fu QP, Sun L, Ao R, Guan XJ, Liu Y, Wang J, He H, Tong WB, Qin ZY, Fan WJ, Li J, He JL, Fang G. 2009. A seroepidemiologic analysis of hepatitis B in Sichuan province. Zhonghua Liu Xing Bing Xue Za Zhi 30:139-143.

Engel LS, Laden F, Andersen A, Strickland PT, Blair A, Needham LL, Barr DB, Wolff MS, Helzlsouer K, Hunter DJ, Lan Q, Cantor KP, Comstock GW, Brock JW, Bush D, Hoover RN, Rothman N. 2007. Polychlorinated biphenyl levels in peripheral blood and non-Hodgkin's lymphoma: a report from three cohorts. Cancer Res 67:5545-5552.

Engels EA, Cho ER, Jee SH. 2010. Hepatitis B virus infection and risk of non-Hodgkin lymphoma in South Korea: a cohort study. Lancet Oncol. 11:827-834.

Environmental Protection Agency. 2008. Hudson River PCBs. http://www.epa.gov/region2/superfund/npl/0202229c.pdf. Accessed 10/19/10.

Fernberg P, Chang ET, Duvefelt K, Hjalgrim H, Eloranta S, Sorensen KM, Porwit A, Humphreys K, Melbye M, Ekstrom Smedby K. 2010. Genetic variation in chromosomal translocation breakpoint and immune function genes and risk of non-Hodgkin lymphoma. Cancer Causes Control 21: 759-769.

Freedman DM, Zahm SH, Dosemeci M. 1997. Residential and occupational exposure to sunlight and mortality from non-Hodgkin's lymphoma: composite (threefold) case-control study. BMJ 314:1451-1455.

Gallagher RP, Macarthur AC, Lee TK, Weber JP, Leblanc A, Mark Elwood J, Borugian M, Abanto Z, Spinelli JJ. Plasma levels of polychlorinated biphenyls and risk of cutaneous malignant melanoma: a preliminary study. Int J Cancer [Epub ahead of print]

Gammon MD, Wolff MS, Neugut AI, Eng SM, Teitelbaum SL, Britton JA, Terry MB, Levin B, Stellman SD, Kabat GC, Hatch M, Senie R, Berkowitz G, Bradlow HL, Garbowski G, Maffeo C, Montalvan P, Kemeny M, Citron M, Schnabel F, Schuss A, Hajdu S, Vinceguerra V, Niguidula N, Ireland K, Santella RM. 2002. Environmental toxins and breast cancer on Long Island. II. Organochlorine compound levels in blood. Cancer Epidemiol Biomarkers Prev 11: 686-697.

Grant WB. 2003. Ecologic studies of solar UV-B radiation and cancer mortality rates. Recent Results Cancer Res 164:371-377.

Gross SA, Zhu X, Bao L, Ryder J, Le A, Chen Y, Wang XQ, Irons RD. 2008. A prospective study of 728 cases of non-Hodgkin lymphoma from a single laboratory in Shanghai, China. Int J Hematol 88:165-173.

Hartge P, Devesa SS, Grauman D, Fears TR, Fraumeni JF, Jr. 1996. Non-Hodgkin's lymphoma and sunlight. J Natl Cancer Inst 88:298-300.

Hartge P, Lim U, Freedman DM, Colt JS, Cerhan JR, Cozen W, Severson RK, Davis S. 2006. Ultraviolet radiation, dietary vitamin D, and risk of non-Hodgkin lymphoma (United States). Cancer Causes Control 17: 1045-1052.

Hayes RB, Yin SN, Dosemeci M, Li GL, Wacholder S, Travis LB, Li CY, Rothman N, Hoover RN, Linet MS. 1997. Benzene and the dose-related incidence of hematologic neoplasms in China. Chinese Academy of Preventive Medicine--National Cancer Institute Benzene Study Group. J Natl Cancer Inst 89:1065-1071.

Hosgood HD 3rd, Purdue MP, Wang SS, Zheng T, Morton LM, Lan Q, Menashe I, Zhang Y, Cerhan JR, Grulich A, Cozen W, Yeager M, Holford TR, Vajdic CM, Davis S, Leaderer B, Kricker A, Schenk M, Zahm SH, Chatterjee N, Chanock SJ, Rothman N, Hartge P, Armstrong B. 2011. A pooled analysis of three studies evaluating genetic variation in innate immunity genes and non-Hodgkin lymphoma. Br J Haemtol [Epub]

Hsu JF, Lee CC, Su HJ, Chen HL, Yang SY, Liao PC. 2009. Evaluation of background persistent organic pollutant levels in human from Taiwan: polychlorinated dibenzo-p-dioxins, dibenzofurans, and biphenyls. Environ Int 35:33-42.

Hu S, Ma F, Collado-Mesa F, Kirsner RS. 2004. Ultraviolet radiation and incidence of non-Hodgkin's lymphoma among Hispanics in the United States. Cancer Epidemiol Biomarkers Prev 13: 59-64.

Hughes AM, Armstrong BK, Vajdic CM, Turner J, Grulich AE, Fritschi L, Milliken S, Kaldor J, Benke G, Kricker A. 2004. Sun exposure may protect against non-Hodgkin lymphoma: a case-control study. Int J Cancer 2004;112: 865-871.

Jacob CO, Fronek Z, Lewis GD, Koo M, Hansen JA, McDevitt HO. 1990. Heritable major histocompatibility complex class II-associated differences in production of tumor necrosis factor alpha: relevance to genetic predisposition to systemic lupus erythematosus. Proc Natl Acad Sci USA 87:1233-1237.

Jaffe E, Harris N, Stein H, Vardiman JW. 2009. Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. 2001. Lyon, IARC Press. World Health Organization Classification of Tumours, Volume 3. Kleihues P and Sobin L.

Jaffe ES, Banks PM, Nathwani B, Said J, Swerdlow SH. 2004. Recommendations for the reporting of lymphoid neoplasms: a report from the Association of Directors of Anatomic and Surgical Pathology. Mod Pathol 17:131-135.

Jin F, Devesa SS, Chow WH, Zheng W, Ji BT, Fraumeni JF, Jr., Gao YT. 1999. Cancer incidence trends in urban Shanghai, 1972-1994: an update. Int J Cancer 83:435-440.

Johnson, K.C., Pan, S., et al., 2003. Residential proximity to industrial plants and non-Hodgkin lymphoma. Epidemiology 14, 687–693.

Jollow DJ, Bruckner JV, McMillan DC, Fisher JW, Hoel DG, Mohr LC. 2009. Trichloroethylene risk assessment: A review and commentary. Crit Reviews in Toxicol 39: 782-797.

Kadin ME, Berard CW, Nanba K, Wakasa H. 1983. Lymphoproliferative diseases in Japan and Western countries: Proceedings of the United States--Japan Seminar, September 6 and 7, 1982, in Seattle, Washington. Hum Pathol 14:745-772.

Lan Q, Shen M, Garcia-Rossi D, Chanock S, Zheng T, Berndt SI, Puri V, Li G, He X, Welch R, Zahm SH, Zhang L, Zhang Y, Smith M, Wang SS, Chiu BC, Linet M, Hayes R, Rothman N, Yeager M. 2007. Genotype frequency and Fst analysis of polymorphisms in immunoregulatory genes in Chinese and Caucasian populations. Immunogenetics 59:839-852.

Lahti TA, Partonen T, Kyyrönen P, Kauppinen T, Pukkala E. Night-time work predisposes to non-Hodgkin lymphoma. 2008. Int J Cancer 123: 2148-2151.

Lan Q, Zhang L, Li G, Vermeulen R, Weinberg RS, Dosemeci M, Rappaport SM, Shen M, Alter BP, Wu Y, Kopp W, Waidyanatha S, Rabkin C, Guo W, Chanock S, Hayes RB, Linet M, Kim S, Yin S, Rothman N, Smith MT. 2004. Hematotoxicity in workers exposed to low levels of benzene. Science 306:1774-1776.

Lan Q, Zheng T, Rothman N, Zhang Y, Wang SS, Shen M, Berndt SI, Zahm SH, Holford TR, Leaderer B, Yeager M, Welch R, Boyle P, Zhang B, Zou K, Zhu Y, Chanock S. 2006. Cytokine polymorphisms in the Th1/Th2 pathway and susceptibility to non-Hodgkin lymphoma. Blood 107: 4101-4108.

Lan Q. Occupational exposure to trichloroethylene and lymphocyte subset toxicity. ICOH meeting. 2009.

Langer P. 2010. The impacts of organochlorines and other persistent pollutants on thyroid and metabolic health. Front Neuroendocrinol. 31: 497-518.

Langford IH, Bentham G, McDonald AL. 1998. Mortality from non-Hodgkin lymphoma and UV exposure in the European Community. Health Place 4: 355-364.

Lee SA, Dai Q, Zheng W, Gao YT, Blair A, Tessari JD, Tian JB, Shu XO. 2007. Association of serum concentration of organochlorine pesticides with dietary intake and other lifestyle factors among urban Chinese women. Environ Int 33:157-163.

Linos, A., Blair, A., et al., 1991. Leukemia and non-Hodgkin’s lymphoma and residential proximity to industrial plants. Arch. Environ. Health 46, 70–74.

Longnecker MP, Rogan WJ, Lucier G. 1997. The Human Health Effects Of DDT (Dichlorodiphenyltrichloroethane) and PCBS (Polychlorinated Biphenyls) and an Overview of Organochlorines in Public Health. Annual Review of Public Health 18: 211-244.

Mandel JH, Kelsh MA, Mink PJ, Alexander DD, Kalmes RM, Weingart M, Yost L, Goodman M. 2006. Occupational trichloroethylene exposure and non-Hodgkin’s lymphoma: a meta-analysis and review. Occup Environ Med 63: 597-607.

McMichael AJ, Giles GG. 1996. Have increases in solar ultraviolet exposure contributed to the rise in incidence of non-Hodgkin's lymphoma? Br J Cancer 73: 945-950.

Michaud DS, Liu Y, Meyer M, Giovannucci E, Joshipura K. 2008. Periodontal disease, tooth loss, and cancer risk in male health professionals: a prospective cohort study. Lancet Oncology 9: 550-558.

Morton LM, Wang SS, Cozen W, Linet MS, Chatterjee N, Davis S, Severson RK, Colt JS, Vasef MA, Rothman N, Blair A, Bernstein L, Cross AJ, De Roos AJ, Engels EA, Hein DW, Hill DA, Kelemen LE, Lim U, Lynch CF, Schenk M, Wacholder S, Ward MH, Hoar ZS, Chanock SJ, Cerhan JR, Hartge P. 2008. Etiologic heterogeneity among non-Hodgkin lymphoma subtypes. Blood 112:5150-5160.

Morton LM, Hartge P, Holford TR, Holly EA, Chiu BC, Vineis P, Stagnaro E, Willett EV, Franceschi S, La Vecchia C, Hughes AM, Cozen W, Davis S, Severson RK, Bernstein L, Mayne ST, Dee FR, Cerhan JR, Zheng T. Cigarette smoking and risk of non-Hodgkin lymphoma: a pooled analysis from the International Lymphoma Epidemiology Consortium (InterLymph). 2005. Cancer Epidemiol Biomarkers Prev 14: 925-933.

Morton LM, Zheng T, Holford TR, Holly EA, Chiu BC, Costantini AS, Stagnaro E, Willett EV, Dal Maso L, Serraino D, Chang ET, Cozen W, Davis S, Severson RK, Bernstein L, Mayne ST, Dee FR, Cerhan JR, Hartge P; InterLymph Consortium. Alcohol consumption and risk of non-Hodgkin lymphoma: a pooled analysis. 2005. Lancet Oncol 6:469-476.

Newton JL, Harney SM, Timms AE, Sims AM, Rockett K, Darke C, Wordsworth BP, Kwiatkowski D, Brown MA. 2004. Dissection of class III major histocompatibility complex haplotypes associated with rheumatoid arthritis. Arthritis Rheum 50:2122-2129.

Ng CS, Chan J, Lo S, Poon Y. 1986. Immunophenotypic analysis of non-Hodgkin's lymphomas in Chinese. A study of 75 cases in Hong Kong. Pathology 18:419-425.

Non-Hodgkin’s Lymphoma Classification Project. 1997. A Clinical Evaluation of the International Lymphoma Study Group Classification of Non-Hodgkin's Lymphoma. Blood 89: 3909-3918

Orsi L, Monnereau A, Dananche B, Berthou C, Fenaux P, Marit G, Soubeyran P, Huguet C, Fenaux P, Marit G, Soubeyran P, Huguet F, Milpied N, Leporrier M, Hemon D, Troussard X, Clavel J. 2010. Occupational exposure to organic solvents and lymphoid neoplasms in men: results of a French case-control study. Occup Environ Med 67: 664-672.

Petridou ET, Dikalioti SK, Skalkidou A, Andrie E, Dessypris N, Trichopoulos D. 2007. Sun exposure, birth weight, and childhood lymphomas: a case control study in Greece. Cancer Causes Control 18:1031-1037.

Porta D, Milani S, Lazzarino AI, Perucci CA, Forastiere F. Systematic review of epidemiological studies on health effects associated with management of solid waste. Env Health 8:60; 2009.

Price P, Bolitho P, Jaye A, Glasson M, Yindom LM, Sirugo G, Chase D, McDermid J, Whittle H. 2003. A Gambian TNF haplotype matches the European HLA-A1,B8,DR3 and Chinese HLA-A33,B58,DR3 haplotypes. Tissue Antigens 62:72-75.

Pronk A, Nuckols JR, De Roos AJ, Airola M, Colt JS, Cerhan JR, Severson R, Blair A, Cleverly D, Ward MH. Residential exposure to industrial combustion facilities and risk of non-Hodgkin lymphoma (EHP, submitted).

Purdue MP, Lan Q, Kricker A, Grulich AE, Vajdic CM, Turner J, Whitby D, Chanock S, Rothman N, Armstrong BK. 2007. Polymorphisms in immune function genes and risk of non-Hodgkin lymphoma: findings from the New South Wales non-Hodgkin Lymphoma Study. Carcinogenesis 28:704-712.

Purdue MP, Engel LS, Langseth H, Needham LL, Anderson A, Barr DB, Blair A, Rothman N, McGlynn KA. 2009. Prediagnostic serum concentrations of organochlorine compounds and risk of testicular germ cell tumors. Environ Health Perspect 117: 1514-1519.

Purdue MP, Bakke B, Stewart P, De Roos AJ, Schenk M, Lynch CF, Bernstein L, Morton LM, Cerhan JR, Severson RK, Cozen W, Davis S, Rothman N, Hartge P, Colt JS. 2010. A case-control study of occupational exposure to trichloroethylene and non-hodgkin lymphoma. Environ Health Perspect 119: 232-238.

Raaschou-Nielsen O, Hansen J, McLaughlin JK, Kolstad H, Christensen JM, Tarone RE, Olsen JH. 2003. Cancer risk among workers at Danish companies using trichloroethylene: a cohort study. Am J Epidemiol 158:1182-1192.

Ramis, R., Vidal, E., et al., 2009. Study of non-Hodgkin’s lymphoma mortality associated with industrial pollution in Spain, using Poisson models. BMC Public Health 9, 26

Rothman N, Cantor KP, Blair A, Bush D, Brock JW, Helzlsouer K, Zahm SH, Needham LL, Pearson GR, Hoover RN, Comstock GW, Strickland PT. 1997. A nested case-control study of non-Hodgkin lymphoma and serum organochlorine residues. Lancet 350:240-244.

Rothman N, Skibola CF, Wang SS, Morgan G, Lan Q, Smith MT, Spinelli JJ, Willett E, De SS, Cocco P, Berndt SI, Brennan P, Brooks-Wilson A, Wacholder S, Becker N, Hartge P, Zheng T, Roman E, Holly EA, Boffetta P, Armstrong B, Cozen W, Linet M, Bosch FX, Ennas MG, Holford TR, Gallagher RP, Rollinson S, Bracci PM, Cerhan JR, Whitby D, Moore PS, Leaderer B, Lai A, Spink C, Davis S, Bosch R, Scarpa A, Zhang Y, Severson RK, Yeager M, Chanock S, Nieters A. 2006c. Genetic variation in TNF and IL10 and risk of non-Hodgkin lymphoma: a report from the InterLymph Consortium. Lancet Oncol 7:27-38.

Scott CS, Chiu WA. 2006. Trichloroethylene cancer epidemiology: a consideration of select issues. Environ Health Perspect 114:1471-1478.

Shen M, Menashe I, Morton LM, Zhang Y, Armstrong B, Wang SS, Lan Q, Hartge P, Purdue MP, Cerhan JR, Grulich A, Cozen W, Yeager M, Holford TR, Vajdic CM, Davis S, Leaderer B, Kricker A, Severson RK, Zahm SH, Chatterjee N, Rothman N, Chanock SJ, Zheng T. Polymorphisms in DNA repair genes and risk of non-Hodgkin lymphoma in a pooled analysis of three studies. 2010. Br J Haematol 151: 239-244.

Skibola CF. Obesity, Diet, and Risk of Non-Hodgkin Lymphoma. 2007. Cancer Epidemiol Biomarkers Prev. 16: 392-395.

Skibola CF, Bracci PM, Halperin E, Conde L, Craig DW, Agana L, Iyadurai K, Becker N, Brooks-Wilson A, Curry JD, Spinelli JJ, Holly EA, Riby J, Zhang L, Nieters A, Smith MT, Brown KM. 2009. Genetic variants at 6p21.33 are associated with susceptibility to follicular lymphoma. Nat Genet 41: 873-875.

Smedby KE, Hjalgrim H, Melbye M, Torrång A, Rostgaard K, Munksgaard L, Adami J, Hansen M, Porwit-MacDonald A, Jensen BA, Roos G, Pedersen BB, Sundström C, Glimelius B, Adami HO. 2005. Ultraviolet radiation exposure and risk of malignant lymphomas. J Natl Cancer Inst 97:199-209.

Soni LK, Hou L, Gapstur SM, Evens AM, Weisenburger DD, Chiu BC. Sun exposure and non-Hodgkin lymphoma: a population-based, case-control study. 2007. Eur J Cancer 43: 2388-2395.

Spinelli JJ, Ng CH, Weber JP, Connors JM, Gascoyne RD, Lai AS, Brooks-Wilson AR, Le ND, Berry BR, Gallagher RP. 2007. Organochlorines and risk of non-Hodgkin lymphoma. Int J Cancer 121:2767-2775.

Stagnaro E, Tumino R, Parodi S, Crosignani P, Fontana A, Masala G, Miligi L, Nanni O, Ramazzotti V, Rodella S, Senoiri Constantini A, Vigano C, Vindigni C, Vineis P. Non-Hodgkin’s Lymphoma and type of tobacco smoke. 2004. Cancer Epidemiol Biomarkers Prev 13: 431-37.

Stellman SD, Djordjevic MV, Muscat JE, Gong L, Bernstein D, Citron ML, White A, Kemeny M, Busch E, Nafziger AN. 1998. Relative abundance of organochlorine pesticides and polychlorinated biphenyls in adipose tissue and serum of women in Long Island, New York. Cancer Epidemiol Biomarkers Prev 6: 489-496.

Stewart P. Organic solvents and risk of non-Hodgkin lymphoma: using a module-based approach to exposure assessment in a case-control study. International Conference on Occupational Health. International Conference on Occupational Health . 2009.

Strachan DP. 2000. Family size, infection and atopy: the first decade of the “ hygiene hypothesis”. Thorax 55 (1 Suppl): S2 – S10.

Strachan DP. 1989. Hay fever, hygiene, and household size. BMJ 299: 1259 – 60.

Swerdlow S, Campo E, Harris NL, Jaffe ES, Pileri SA, Stein H, Thiele J, Vardiman JW. 2009. World Health Organization Classification of Tumours of Haematopoietic and Lymphoid Tissues. IARC Press: Lyon 2008.

Toft G, Hagmar L, Giwercman A, Bonde JP. 2004. Epidemiological evidence on reproductive effects of persistent organochlorines in humans. Reprod Toxicol. 19: 5-26.

Turner JJ, Hughes AM, Kricker A, Milliken S, Grulich A, Kaldor J, Armstrong B. 2004. Use of the WHO lymphoma classification in a population-based epidemiological study. Ann Oncol 15:631-637.

Vajdic CM, Falster MO, de Sanjose S, Martínez-Maza O, Becker N, Bracci PM, Melbye M, Smedby KE, Engels EA, Turner J, Vineis P, Costantini AS, Holly EA, Kane E, Spinelli JJ, La Vecchia C, Zheng T, Chiu BC, Dal Maso L, Cocco P, Maynadié M, Foretova L, Staines A, Brennan P, Davis S, Severson R, Cerhan JR, Breen EC, Birmann B, Cozen W, Grulich AE. 2009. Atopic disease and risk of non-Hodgkin lymphoma: An Interlymph pooled analysis. Cancer Res 69:6482-6489.

Vlaanderen J, Lan Q, Kromhout H, Rothman N, Vermuelen R. 2010. Occupational benzene exposure and the risk of lymphoma subtypes: a meta-analysis of cohort studies incorporating three study quality dimensions. Environ Health Perspect [Epub ahead of print].

Wang SS, Cerhan JR, Hartge P, Davis S, Cozen W, Severson RK, Chatterjee N, Yeager M, Chanock SJ, Rothman N. 2006. Common genetic variants in proinflammatory and other immunoregulatory genes and risk for non-Hodgkin lymphoma. Cancer Res 66:9771-9780.

Wang SS, Purdue MP, Cerhan JR, Zheng T, Menashe I, Armstrong BK, Lan Q, Hartge P, Kricker A, Zhang Y, Morton LM, Vajdic CM, Holford TR, Severson RK, Grulich A, Leaderer BP, Davis S, Cozen W, Yeager M, Chanock SJ, Chatterjee N, 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: e5360.

Wang SS, Cozen W, Cerhan JR, Colt JS, Morton LM, Engels EA, Davis S, Severson RK, Rothman N, Chanock SJ, Hartge P. 2007. Immune mechanisms in non-Hodgkin lymphoma: joint effects of the TNF G308A and IL T3575A polymorphisms with non-Hodgkin lymphoma risk factors. Cancer Res 67: 5042-5054.

Wang SS, Davis S, Cerhan JR, Hartge P, Severson RK, Cozen W, Lan Q, Welch R, Chanock SJ, Rothman N. Polymorphisms in oxidative stress genes and risk of non-Hodgkin lymphoma. 2006. Carcinogenesis 27: 1828-1834.

Wartenberg D, Reyner D, Scott CS. 2000. Trichloroethylene and cancer: epidemiologic evidence. Environ Health Perspect 108 Suppl 2:161-176.

Weihkopf T, Becker N, Nieters A, Mester B, Deeg E, Elsner G, Blettner M, Seidler A. 2007. Sun exposure and malignant lymphoma: a population-based case-control study in Germany. Int J Cancer 120:2445-2451.

Willis-Karp M, Santeliz J, Karp CL. 2001. The germless theory of allergic disease: revisiting the hygiene hypothesis. Nat Rev Immunol 1:69-75.

Wong KY, Chia SE, Kuperan P et al. Sun exposure and the risk of malignant lymphoma in an

Asian population: The Singapore Lymphoma Study (Abtract #1824). In: Proceedings of the 101^st

Annual Meeting of the American Association for Cancer Research, 2010.

Zhang J, Mauzerall DL, Zhu T, Liang S, Ezzati M, Remais JV. 2010. Environmental health in China: progress towards clean air and safe water. Lancet 375:1110-1119



45