Introduction

While the incidence of non-Hodgkin's lymphoma (NHL) has doubled over the past two decades in the U.S., the etiology of the most common types of lymphoma remains elusive (American Cancer Society, 2001). The treatment and, ultimately, the prevention of this disease are dependent upon our success in identifying those factors that incite the emergence and promotion of these hematologic malignancies. While some factors associated with an increased risk of this disease have been identified, the elucidation of clear etiologic factors and their mechanistic role in pathogenesis has been challenging.

The lymphomas encompass an array of heterogeneous malignancies that originate in lymphocytes. The current taxonomy, recently adopted by the World Health Organization (Harris et al., 1999) for the classification of the various types of lymphoma, represents a revision of the previous system known as the Revised European–American Classification of Lymphoid Neoplasms (REAL) (Harris et al., 1994). This classification recognizes Hodgkin's disease, B-cell neoplasms, and T/NK-cell tumors. Hodgkin's Disease, distinguishable by the presence of multi-nucleated Reed–Sternberg cells (Reed, 1902), represents only about 12% of all lymphomas. This neoplasm is somewhat unique related to its epidemiology, etiologic links, treatment, and prognosis, thus it will not be discussed further. This review will focus on the more common non-Hodgkin's lymphomas which encompass a large number of distinct entities that are recognized based on their natural history as well as morphologic, immunophenotypic, and genetic features.

Lymphocytes arise from hematopoietic stem cells and undergo differentiation to a specific phenotype expressing unique cell surface receptors and having a distinct morphology. Lymphoma tumor cells are a malignant form of these precursor lymphocytes arrested at a specific stage of differentiation. The transformation of these cells represents a multi-step process involving the progressive accumulation of genetic lesions that result in the clonal expansion of malignant T- or B-cells, and the establishment of either a solid or leukemic malignant process. Chromosomal translocations, usually balanced reciprocal recombinations, are the genetic hallmark of lymphoid malignancies; their presence has been confirmed in up to 90% of NHL cases (Offit et al., 1991; Grogan and Miller, 1995; Ye, 2002). Many of these aberrations represent errors that occur during antigen receptor gene rearrangements involving the immunoglobulin genes of B-cells and the T-cell receptor genes (Tycko and Sklar, 1990). At a molecular level, these translocations with or without additional chromosomal deletions and mutations may precipitate oncogene activation or tumor suppressor gene inactivation. Oncogenic viruses provide another potential avenue for genetic lesions in NHL; these viruses are capable of altering the lymphoid cell genome by the introduction of exogenous genes that interfere with normal cell growth and regulation (Gaidano and Dalla-Favera, 1997).

The diversity of lymphoma subtypes has introduced challenges in the investigation of this complex disease; researchers constantly struggle with consideration of lymphoma as a single disorder versus the more likely possibility that it actually represents a mixture of different disease entities with potentially varying etiologies.

Epidemiology

Descriptive epidemiology

In 2004, 54 370 new cases of NHL and approximately 19 410 deaths from the disease are expected (American Cancer Society, 2004). As the fifth most commonly occurring cancer in both men and women, there are approximately 19 new cases per 100 000 persons in the US each year (Greenlee et al., 2001; SEER, 2003). The disease occurs more often in whites than in blacks, and it is about 50% more common among men than women. Between 1978 and 1995, the annual age-adjusted incidence rates for white and black males were 17.1 and 11.5 per 100 000, respectively. For white and black women, the rates were 12.6 and 7.4 per 100 000 (Groves et al., 2000). Incidence rises exponentially with age, as shown in Figure 1a and b. In persons over the age of 65, the incidence of disease rapidly increases to 68/100 000. While aggressive lymphomas are the most common lymphoid neoplasm in young adults, both aggressive and indolent lymphoid malignancies are more typically diagnosed in patients 50 years of age and older.

Figure 1.

(a) SEER 1996–2000 age-specific incidence rates of NHL by gender; (b) SEER 1996–2000 age-specific incidence rates of NHL by race

Full figure and legend (139K)

NHL occurs worldwide and is strongly correlated with socioeconomic status. The disease is most common in developed countries, with US rates exceeding those of most other countries in the world. Notable are the unusually low rates of NHL reported in Thailand and China (2–3/100 000). The occurrence of NHL subtypes varies by geographic area. For instance, Burkitt's lymphoma is most common in tropical Africa; the highest rates of adult T-cell lymphoma occur in southwest Japan and the Caribbean; and follicular lymphomas are extremely rare in Latin America (Shipp et al., 1997).

Therapeutic approaches for this disease are based on the specific lymphoma subtype, the stage of disease, the physiologic status of the patient, and the prognosis. While chemotherapy and/or radiotherapy are curative in some patients, many with primary or relapsed disease remain refractory to conventional treatments. For some subtypes, extended remissions have been achieved using either newly developed monoclonal antibodies directed against lymphoma cells or high-dose chemotherapy regimens paired with stem cell transplantation. Survival varies considerably by specific subtype; however, the overall 5-year survival rate for this malignancy remains relative stable at 53% (Holford et al., 1992; Ries et al., 2001).

Trends in NHL incidence

Non-Hodgkin's lymphoma rates have steadily increased 3–4% each year in the US from 1973 to the mid 1990s (refer to Figure 2). Most other Westernized countries in which registry data are available have reported similar trends (Devesa and Fears, 1992). These increases in NHL have been observed in males and females, whites and nonwhites, and all age groups except the very young. With the exception of skin malignancies, such temporal increases in cancer incidence are unprecedented.

Figure 2.

SEER 1975–2000 age-adjusted incidence rates of NHL by gender

Full figure and legend (59K)

Since 1980, NHL incidence in males, aged 25–54, has undergone dramatic escalation, mostly related to the HIV epidemic. While high-grade lymphomas have doubled in females and tripled in males between 1978–83 and 1990–95, a larger absolute increase has occurred in the intermediate grade lymphomas (Groves et al., 2000). While increases in disease incidence have been documented in essentially every subtype of NHL, patterns of disease at presentation have also changed. In the past two decades, there has been a sharp rise in the incidence of extranodal primary lymphomas, representing 33% of all incident lymphomas in the US in 2000. Although stomach and skin remain the most common extranodal sites, primary disease in the brain has sharply increased (Groves et al., 2000).

Improved cancer reporting, more sensitive diagnostic techniques particularly for borderline lesions, changes in classification of lymphoproliferative diseases, and, in particular, the increasing occurrence of AIDS-associated lymphomas have contributed to the startling escalation of disease incidence (Hartge and Devesa, 1992). However, extensive analyses have led to the conclusion that these factors account for only about 50% of the additional cases of NHL (Banks, 1992; Hartge and Devesa, 1992; Holford et al., 1992). Since 1995 NHL incidence among males has decreased somewhat, reflecting the decline in the number of HIV positive individuals developing AIDS and the improved immune status of those with AIDS resulting from the introduction of highly active anti-retroviral therapy (HAART). However, available data suggest that the non-AIDS-related NHL incidence rates have continued to increase, specifically the rates among females, older males, and blacks (Eltom et al., 2002).

Etiology of NHL

Pathogenesis

The biologic mechanisms responsible for the initiation of cancer and the continuous promotion of tumor growth through proliferation and expansion of a malignant cell remain a complex mystery. However, it seems certain that the accumulation of genetic lesions may render cells malignant, and clonal expansion of these cells is allowed due to dysregulation of cell growth, cell signaling pathways, programmed cell death, and/or immune eradication processes. Although the specific mechanism responsible for the development of NHL has not been determined, a few components of the process have been suggested. Chronic antigenic stimulation increases B-cell proliferation, which in turn increases the probability of a random genetic mistake, particularly related to immunoglobulin gene rearrangements. Factors that induce proliferation thereby potentially lead to more errors. In cases in which a virus acts as the foreign stimulant, the virus itself may infect a normal cell and integrate viral DNA into the host genome, thereby transforming the cell into a malignant cell capable of self-replication. In either case, antigenic stimulation has been demonstrated to lead to a compensatory downregulation of the T-cell response, that is, an immunosuppressive state. Other factors that independently precipitate an immunosuppressive state may act as co-factors in lymphomagenesis by further inhibiting the recognition and protective eradication of a malignant cell. Although the data are less compelling, another possible mechanism is that a lymphoid cell undergoes mutation by an environmental carcinogen and its growth progresses uninterrupted due to some immune dysfunction.

The uniform escalation in NHL incidence, independent of both demography and geography, implicates an increasing exposure either in time or intensity to a ubiquitous lymphomagenic agent. Although a myriad of endogenous and exogenous factors have been investigated, the etiology of most types of NHL remains elusive.

Hereditary factors

While cancer is considered a disease of the genes, most genetic lesions linked to other types of cancer represent somatic mutations. However, since evidence suggests that oncogenes, tumor suppressor genes, and DNA repair genes play an important role in carcinogenesis, inherited (germline) genetic susceptibility to NHL is conceivable, and certainly deserves consideration. Although unusual, aggregation of NHL in certain families has been observed (Linet and Pottern, 1992; Holly et al., 1999). Having close relatives with NHL or other lymphoproliferative diseases has been associated with a 2.5–4-fold increase in the risk of NHL (Zhu et al., 1998; Paltiel et al., 2000). Concerns have been raised that this association may merely represent the early identification of new cases in families with an increased awareness/knowledge of the early symptoms of NHL (Wiernik et al., 2000). It is not clear whether familial clustering is truly attributable to genetic predisposition or rather to similar environmental exposures.

Primary, congenital immunodeficiency has been associated with a 12–25% risk of malignancy, of which NHL is predominant (Filipovich et al., 1992). Ataxia telangiectasia, a condition marked by DNA processing and repair defects and impaired cell-mediated immunity, is associated with a 10% risk of diffuse large B-cell lymphomas. Approximately 15% of individuals with Wiscott–Aldridge syndrome develop diffuse large B-cell lymphomas; these neoplasms are primarily extranodal, frequently involve the central nervous system (CNS), and are usually Epstein-Barr virus (EBV) positive (Levine, 1994). Both subacute combined immunodeficiency and X-linked lymphoproliferative syndromes are also associated with a greater risk of aggressive lymphomas, often EBV positive and presenting in an extranodal site (Filipovich et al., 1992).

In general, familial lymphomas account for less than 5% of all cases, and primary immunodeficiency is quite rare; therefore, hereditary factors are unlikely to account for a substantial number of cases, and they are unlikely to explain the temporal trends in NHL incidence in the last quarter of the 20th century.

Immunosuppression

The only predisposing factor both consistently and strongly associated with NHL risk is immunosuppression (Filipovich et al., 1992; Kinlen, 1992). Although the incidence of most common forms of cancer among immunosuppressed individuals does not differ from that of the normal population, the data clearly suggest an important role of the immune system in the occurrence of NHL. Three host factors have been implicated in this increased risk: a high prevalence of EBV, defects in immunoregulation with resultant cytokine production, and genetic defects precipitating abnormal immunoglobulin and T-cell receptor gene rearrangement during lymphopoiesis (Filipovich et al., 1992). Both the congenital (as previously discussed) and acquired immunodeficiencies that predispose persons to NHL have common features of immune dysregulation and persistent, chronic antigenic stimulation (Hoover, 1992). In this setting, aggressive or highly aggressive B-cell tumors often develop that are widely disseminated and are frequently associated with EBV (Filipovich et al., 1992).

As with congenital immunodeficiencies, increases in NHL are seen with autoimmune conditions such as rheumatoid arthritis (Kamel et al., 1995), systemic lupus erythematosus (Pettersson et al., 1992), Sjogren's syndrome, and celiac disease (Harris et al., 1967). Although immunosuppressive drugs used in the treatment of these conditions may cause an increase in NHL incidence, evidence suggests that the persistent inflammatory activity of the autoimmune process may have a direct link with increased risk of lymphomagenesis (Jonsson et al., 1999). Wahl et al. (1983) demonstrated that these conditions may be accompanied by impaired T-cell function, which may interfere with an immune response to viruses and emerging malignant cells. Baecklund et al. (1998) demonstrated a 25-fold increase in the risk of NHL among persons with highly inflammatory rheumatoid arthritis as compared to a similar group having low inflammatory disease; this risk was independent of treatment.

Individuals with secondary immunodeficiency, both drug-induced for the purpose of either suppressing post-transplant organ rejection or treating autoimmune diseases and acquired as a result of HIV infection, also have a striking risk of NHL. Many of these malignancies are associated with the presence of EBV infection. In this population, tumors are characterized by short latency, high grade, and a proclivity for extranodal sites (Cleary and Sklar, 1984; Kinlen, 1992). In organ transplant populations, the magnitude of the excess risk of NHL varies widely depending on the degree, duration, and type of immunosuppression, but relative risks approach 67.0 (Opelz and Henderson, 1993) Cardiac transplant patients, considered to require the highest doses of immunosuppressive drugs, develop NHL as the most commonly occurring malignancy. The cumulative incidence rate of NHL among cardiac transplant recipients approaches 5%. Swinnen et al. (1990) reported a substantial increase in NHL incidence among cardiac transplant patients treated with the monoclonal antibody OKT3, which profoundly depletes circulating CD3+ T lymphocytes. A striking feature of post-transplant lymphomas is the short interval from transplant to clinical presentation of malignancy, often within 6 months, and a remarkable predilection for the brain (Palackdharry, 1994). Some of these tumors reflect an outgrowth of EBV-immortalized cells, suggesting a failed immunosurveillance (Ambinder et al., 1999). Primary EBV infection after transplant has been identified as a powerful risk factor in this population, accounting for the particularly high rate of post-transplant lymphomas in children who were EBV negative at the time of transplant. The critical role of immunosuppression in the development of these lymphomas is evidenced by the high rate of tumor regression observed in those cases in which immunosuppressive drugs are withdrawn.

Acquired immunodeficiency syndrome (AIDS)

NHL represents one of the most common malignancies associated with human immunodeficiency virus (HIV) infection. As early as 1985, NHL was recognized as an AIDS-defining condition; approximately 4% of all AIDS cases first present with NHL (Selik et al., 1987; Del Maso and Franceschi, 2003). The estimated relative risk of NHL among those with AIDS is 150–250 (Goedert, 2000). These tumors are invariably derived from B-cells and most often present as either high (60%) or intermediate grade malignancies. With 60–70% being stage III or IV at presentation, these neoplasms are characterized by a widespread extent of disease at diagnosis and usually occur late in the course of AIDS. HIV is not considered to be an oncogenic virus, and, in fact, HIV sequences have never been unequivocally demonstrated within neoplastic lymphoid cells. HIV may, however, contribute to lymphomagenesis by impairment of cell-mediated immunity and increased opportunity for virally induced cell proliferation, leading to an accumulation of genetic lesions (IARC Working Group, 1996). HIV-associated cancer risk increases as CD4 counts decrease, suggesting that increasing immunosuppression in the presence of viral antigenic stimulation may be critical in lymphomagenesis (Biggar et al., 2001). Approximately 80% of AIDS-related NHLs occur at extranodal sites, of which the CNS is the most common. EBV is consistently detected in these tumors as well as in the cerebrospinal fluid of AIDS-related cases (Cinque et al., 1993). The AIDS epidemic has had no measurable effect on the incidence of low-grade NHL (Rabkin et al., 1991).

Although the number of persons being diagnosed with HIV infection continues to increase, median survival has also increased. HIV-related mortality has decreased by 50% in the US (Palella et al., 1998). This extended survival is attributable to HAART, which is now used to treat 60% of all HIV-positive individuals in the US (Jacobson et al., 1999; Moore and Chiasson, 1999). Improved survival with increasing disease incidence results in a rapidly inflated prevalence rate of disease. Sustained elevations of CD4 lymphocyte counts and effective suppression of HIV replication in some subjects treated with HAART have led to partial immune restoration, which is perceived to have resulted in a 42% reduction in AIDS-related NHL in 1997–99 compared to 1992–96. This decrease, however, has only been seen with CNS lymphomas; there has been minimal to no reduction in systemic NHLs. Scientists worry that this recent reduction in incident cases of lymphoma may represent a mere delay in the onset of AIDS-related malignancy, and eventually greater rates of NHL will be observed as individuals with AIDS survive longer. Clearly, the long-term repercussions of HAART on AIDS-related NHL incidence have yet to be defined. In the meantime, the challenge to understand the interaction between significant HIV-induced immunodeficiency, virally induced antigenic stimulation related to HIV, EBV, and/or HHV8, and resulting B-cell proliferation in the initiation and promotion of NHL continues.

Ultraviolet radiation

While there is no consistent evidence linking ionizing radiation to NHL, ultraviolet radiation (UVR) has been associated with NHL. Increasing recreation time, changing fashions leading to more skin exposure, and the cosmetic desirability of tanning have contributed to increasing UVR exposure in recent years. The increasing incidence of NHL parallels the escalating annual rates of skin cancer in the US, with high rates of disease in sunny geographic areas. Such disease pairing may sometimes provide etiologic clues to disease causation. The immune system appears to play an important role in both skin cancer and NHL. Among immunosuppressed organ transplant recipients, a 50–100-fold increase in skin cancer risk has been reported. Levi et al. (1996) confirmed an excess of skin cancers following NHL and an excess of NHL following skin cancer, again suggesting a common etiologic factor.

Photoimmunology, the study of the immunological effects of UVR, is now recognized as a unique entity of scientific investigation. About 25 years ago, the pioneering work by Fisher and Kripke (1982) demonstrated in an animal model the profound effect of UVR in not only changing normal cells into cancer cells but also permitting the outgrowth of transformed cells by depressing the immune response (Vink et al., 1996). The immune alteration appeared to be systemic, since tumors implanted outside of an irradiated area grew progressively. The decrease in immune function was specifically targeted in studies in which chronic exposure of mouse T-lymphocytes to UVB radiation caused inhibition of immune responsiveness to antigenic tumors and also interfered with both contact and delayed hypersensitivity reactions. Kripke and colleagues demonstrated the critical role of soluble mediators in this response by showing that exposure of murine keratinocytes to UV radiation in vitro caused the production of soluble factors that mimicked whole body UV irradiation (Fisher and Kripke, 1982). Injection of these factors into normal mice, followed by immunization, induced active suppression. IL-10 and TNF alpha were identified as major mediators of suppression in delayed and contact hypersensitivity reactions, respectively (Vermeer and Streilein, 1990; Rivas and Ullrich, 1992).

In humans, sunlight exposure in susceptible individuals produces acute UV injury, resulting in both inflammatory and immunomodulatory responses; these represent two integrated, complex processes. Kripke (1994) suggests that it is the absorption of UVR by these cells and resulting DNA damage that triggers the biochemical cascade of immunosuppression. Responses include a transient decrease in human NK-cell number and activity, and a decrease in helper T-cell subsets, accompanied by an increase in suppressor T-cell subsets. Studies have demonstrated that such UV irradiation can also impair immunity to infectious agents (Jeevan and Kripke, 1990). Such findings suggest that increased sun exposure may alter the balance of the host–disease relationship in favor of pathogens.

Studies have implicated UVR in risk of NHL, but data remain inconclusive (Zheng et al., 1992; Melbye et al., 1996). A moderate international geographic correlation of UVR levels and NHL incidence has been reported (McMichael and Giles, 1996). Investigations of the cancer-causing effect of Agent Orange demonstrated the highest rates of NHL among veterans stationed on the ground in low exposure areas and sailors off the coast (Breslin et al., 1988). These results allow for consideration of an alternate hypothesis, that is perhaps increased exposure to UVR with resultant subclinical immunosuppression could account for the greater than expected number of NHL occurrences in this study sample. Similarly, findings of a higher risk of NHL among farmers may be attributed to the increased UVR exposure that accompanies outdoor work as opposed to chemical exposures, which are more often examined in these studies.

Viruses and other pathogens

Recently, the number of viruses and microbes that have been associated with HIV-negative NHL is steadily increasing. The role of these pathogens in NHL will briefly be discussed.

EBV

A member of the herpes family, EBV, is a ubiquitous virus that persistently infects approximately 95% of adults in the US. While primary infection is usually subclinical in nature, the virus is maintained indefinitely as a plasmid in the cytoplasm of resting B-cells. While it appears that the majority of NHLs in persons without primary or secondary immunodeficiency are EBV-negative, viral prevalence has not been systematically examined to optimally describe the epidemiology of EBV in this population.

In Africa and New Guinea, Burkitt's lymphoma is endemic and evidence has consistently demonstrated episomal EBV genome in the tumor tissue, whereas less than 20% of Burkitt's lymphomas occurring in other geographic locales are EBV positive. In the normal host, EBV-driven lymphoproliferation is thought to be primarily controlled by EBV-specific autologous cytotoxic T-cells, antibody-dependent cellular cytotoxicity, natural killer cell activity, and endogenous interferon. The EBV-specific immunological control is delicately balanced to maintain the EBV in latency following primary infection. In the immunodeficient host, suppressor and cytotoxic functions are often defective and EBV-infected B cells can resume proliferation. The high mitotic index of EBV-transformed cells increases the probability of mutational events that precipitate the growth of cells that bear cytogenetic rearrangement and that are not susceptible to regulatory stimuli (Filipovich et al., 1992). It appears that, among immunocompetent individuals, the immune system does not protect against the initial EBV infection; however, the development of malignant clones of infected B-cells is prevented by the immune system (Kast et al., 1998). It has been proposed that this process of immunosurveillance may be dependent upon the recognition of viral antigens on the tumor cell surface by cells of the immune system, thus explaining the high rate of EBV-associated NHL observed in immunosuppressed populations. Unlike the endemic forms of Burkitt's lymphoma in which episomal EBV is found in the genome of tumor cells, the sporadic types are indistinguishable by histopathology and chromosomal abnormalities, yet 10–20% are EBV positive. EBV transforms lymphocytes and produces polyclonal tumors in subhuman primates and immunosuppressed patients, indicative of its oncogenic potential. Since almost 90% of people in developed countries are infected with EBV, as are 99% of those in developing countries, it is clear that the virus may be necessary but not sufficient for lymphomagenesis. Immunosuppression due to widespread malaria in Africa may potentiate EBV effects (Morrow et al., 1976). It has been suggested that as EBV genome-positive cells increase during malaria infection, so does the chance of cytogenetic abnormalities occurring in B cells with potential development of Burkitt's lymphoma. The actual role of EBV in maintaining the transformed phenotype is not clear.

HTLV

Adult T-cell leukemia/lymphoma (ATLL) was first identified in 1976 by Takatsuki et al. (1976) and described in a cluster of southern islands in Japan. In 1978, human T-cell lymphotropic virus-1 (HTLV-1) was identified as the etiologic agent, a retrovirus producing slow, progressive infection and known to precipitate immunosuppression. The evidence that HTLV-1 has an etiologic role in ATLL is compelling, as the malignant cells in essentially all antibody-positive cases have monoclonal integrated provirus (Mueller et al., 1992). While the mechanism of pathogenesis is not clear, it has been proposed that repeated virus replication leads to polyclonal expansion of the HTLV-1-infected activated T helper cells. In healthy individuals, the expansion of these cells is controlled by immunity against viral proteins, but occasionally repeated replication of an infected cell occurs. Such cells then undergo monoclonal expansion and may progress to a malignancy (Yosida et al., 1985).

This virus is highly endemic in Japan and the Carribean; however, cases have been identified among previous residents of viral endemic regions. In these regions, ATLL accounts for 50% of all lymphoid malignancies. Exposure to the virus early in life is not uncommon; such early transmission is associated with the greatest risk of disease (Cleghorn et al., 1995). Carriers of HTLV-1 have a cumulative lifetime risk of ATLL of 1–5% (Murphy et al., 1991).

Helicobacter pylori

Helicobacter pylori, a gastric pathogen, causes chronic gastritis and has been shown to have a pivotal role in peptic ulcer disease and gastric lymphoma particularly of the MALT (mucous-associated lymphoid tissue) type. Wotherspoon et al. (1991) found H. pylori infection in 92% of the cases of primary gastric MALT lymphoma, and suggested that infection-induced gastritis provides the background lymphoid tissue in which NHL develops. A sixfold increase in risk (CI 2.0–19.9) of gastric lymphoma related to H. pylori seropositivity has been demonstrated with no increase in lymphomas at other sites (Parsonnet et al., 1994). In vitro experiments have demonstrated that malignant B cells proliferate only after T-cell specific activation by H. pylori (Hussell et al., 1996). The association is now well established based on: evidence demonstrating an association between previous H. pylori infection and the development of primary gastric lymphoma, confirmation of the importance of H. pylori in the pathogenesis of gastric lymphoma in vitro studies (Hussell et al., 1993), and clinical evaluation of the positive effects of eradicating the organism in cases of low-grade lymphoma. The lymphoproliferative conditions have shown regression with HP-specific antibiotic therapy (Wotherspoon et al., 1991). These findings support that chronic antigenic stimulation and inflammation may precipitate this lymphoid malignancy.

Human herpes virus 8

Human herpes virus 8 (HHV8), primarily associated with Kaposi's sarcoma, has been detected in the majority of subjects with primary effusion lymphoma, a rare presentation of a body cavity-based lymphoma seen almost exclusively in HIV positive patients. These patients often present with dual EBV and HHV8 infection; therefore, delineation of the etiologic role of each virus is difficult (IARC, 1997). These neoplasms have been demonstrated to be monoclonal expansions of a single infected cell, suggesting that the viral infection precedes tumor growth (Melbye and Trichopoulos, 2002).

Hepatitis C

Approximately 170 million people worldwide are infected with Hepatitis C virus (HCV), a single-strand RNA virus. Although not known to be oncogenic, HCV is thought to have some immune modulatory effects. HCV is lymphotropic and has been shown to replicate in peripheral blood mononuclear cells. Several investigators have reported an increase risk of NHL with hepatitis C infection (Luppi et al., 1996; De Vita et al., 1997) A recent meta analysis of 48 studies of B-cell NHL (5542 patients) showed a HCV prevalence among these cancer patients of 13% (CI 12–14%). In 10 of these studies in which matched controls were included, the background prevalence was 1.5%. The association between HCV and NHL is strongest in geographic areas with the highest prevalence of the viral infection. HCV prevalence in the US is lower than many European countries; in US studies the rate of HCV in NHL cases is reported to range from 0 to 22%. It has been proposed that the E2 protein of HCV may be responsible for chronic antigen-driven polyclonal B-cell proliferation, which may lead to lymphomagenesis (De Re et al., 2000; Gisbert et al., 2003).

Simian virus 40

Simian virus 40 (SV40), a monkey polyomavirus known to contaminate the Salk polio vaccine, has been demonstrated to cause tumors in laboratory animals. Some of these tumors were identified as lymphoid malignancies. Two early investigations of SV40 and NHL were conducted (Rizzo et al., 1999; David et al., 2001). Although in both studies SV40-positive specimens were identified, viral copy numbers were quite low. Both studies concluded that there was no evidence of an association between SV40 and NHL. In contrast, two recent reports convincingly showed SV40 sequences in NHL (Shivapurkar et al., 2002; Vilchez et al., 2002). Both found nearly identical rates of SV40 sequences (42 and 43%). Both diffuse and follicular types of B-cell lymphomas, the most common types, were positive. There was no difference in rates of SV40 positivity between HIV-positive and HIV-negative lymphomas. Five of the SV40-positive NHL tumors were from subjects born after 1963, suggesting that persons not inoculated with contaminated polio vaccine may, in fact, have been infected with SV40. There is evidence that SV40 is transmitted both vertically and horizontally. Serologic surveys suggest infection rates of 3–20% among recipients of contaminated as well as noncontaminated vaccine. A recent retrospective study in Japan demonstrated a fourfold increase in the odds of SV40 in large B-cell lymphomas (19%) compared to controls (4.7%), suggesting that SV40 may be a candidate etiologic factor for lymphoma in Japan as well as in the US (Nakatsuka et al., 2003). Laboratory studies suggest that SV40 is capable of transforming human cells. While the oncogenic potential of SV40 is attributed to large T antigen, a multifunctional transforming protein that is required for DNA replication, the direct role of this virus in the development of NHL is not clear. Given the high prevalence of the virus reported in lymphoma specimens tested in some studies, further investigation of this infectious agent is warranted.

Other pathogens

Recently, infection with Borrelia burgdorferri has been linked to lymphoproliferative conditions, which have been known to evolve into a primary cutaneous B-cell lymphoma. The majority of these observations have come from European countries, with little evidence of such an association in North America (Willemze et al., 1997). Mediterranean lymphoma is an unusual form of lymphoma that arises in small intestinal mucosa-associated lymphoid tissue. Early-stage disease regresses with antibiotic treatment, suggesting a bacterial etiology. In a recent small series, five of seven tumor specimens of Mediterranean lymphoma tested positive for the bacteria, Campylobacter jejunei (Lecuit et al., 2004). Recently, an association between Chlamydia psittaci infection and ocular adnexal lymphoma was also reported (Ferreri et al., 2003). NHL has also been associated with a history of numerous chronic infectious diseases including herpes zoster, pyelonephritis, tuberculosis, and malaria (Tavani et al., 2000). Clearly, the number of pathogens that may be responsible for immunoproliferative neoplasms is growing. The specific role of these infectious agents may relate to the initiation of the malignancy by immunostimulation or to disease promotion via immunosuppression.

Summary

As medical care continues to improve and more potentially immunocompromised individuals enjoy increased longevity, an increased incidence of NHL should be anticipated. Further advances in molecular diagnostic techniques may allow additional viruses to be identified that may have a role in the pathogenesis of lymphoma. Regardless of the infecting organism, it appears that in humans immunosuppression is a critical phase during which latent infections precipitate tumorigenic processes.

Chemical/agricultural exposures

Given the excess mortality from cancer observed in the central US from 1950–1980, agriculture-related environmental exposures have been ambitiously investigated. The role of environmental exposures in NHL is uncertain since the studies are marked by small study populations with poorly characterized exposures. Problems include: limited statistical power; poor quality of exposure measurements; minimal efforts to control for confounding; and lack of evidence related to dose response, temporality, and biological plausibility. Some studies have demonstrated high rates of NHL among agricultural workers and persons working with pesticides. Death certificate analysis showed that while farmers had a lower risk of cancer overall, they had higher than expected rates of lymphoma (Blair and Zahm, 1991). While many studies were positive, they did not specifically address the causative exposure which in farmers could include pesticides, animal pathogens, and UVR.

Chemical exposures have been shown to increase genetic mutations and alter cell-mediated immunity (D'Amore et al., 1992). These substances may, therefore, generate a genetic mutation followed by a systemic immunosuppressive effect, which may interfere with normal immunosurveillance. Exposure classification for many of the studies related to chemical carcinogens is frequently based on self-report from known lymphoma cases; therefore, risk of bias must be considered. Phenoxy herbicide, which is widely utilized both in agriculture and in the general population, has been most consistently associated with higher risk of disease; however, inadequate evidence of a dose–response relationship has raised doubts regarding the significance of this exposure (Pearce, 1989; D'Amore et al., 1992). Herbicide 2,4-D has been associated with a 50–200% excess of NHL, with the most frequent or heaviest exposure generating relative risks of 3–8.0 (Hoar et al., 1986; Wigle et al., 1990; Zahm et al., 1990). However, in a recent review, Garabrant concludes that there is scant epidemiologic evidence to support any increase in cancer of any type due to 2,4-D exposure (Garabrant and Philbert, 2002). Canine lymphoma has similarly been associated with dog owners who use 2,4-D and/or commercial lawn pesticide treatments. Lawn pesticide use is increasing 5–8% per year in the US and applications are thought to be completed five times more frequently than the application rate that is used on farms (Pimental et al., 1993). Insecticides have also been linked to NHL (Hayes et al., 1991). Additional studies of specific occupations and specific chemicals have suggested some increase in the risk of NHL among exposed persons; however, exposure assessment is quite difficult due to changing patterns over time. These problems of exposure measurement/classification have limited the conclusion that can be drawn from these studies.

Hair dyes

Hair dyes contain compounds that are mutagenic and carcinogenic in animals (IARC, 1978). Excess lymphatic cancers have been reported in both hairdressers and persons whose hair is dyed. A recent study by Zhang et al. (2004) showed a 30% increase in NHL among women coloring their hair before 1980. In 1979, the FDA required a cancer warning label on hair dyes, which led to reformulation of all oxidative dye products. The highest risk was noted with dark hair dye (RR 2.1 (CI 1.0–4.0)) for more than 25 years of use. The greatest risk was in low-grade B-cell tumors. Although increasing numbers of women are coloring their hair, hair dye is unlikely to account for a significant proportion of the recent increases in NHL incidence, particularly among men.

Tobacco and alcohol

Tobacco has been shown to alter the immune response, and contains substances known to be leukemogenic; however, minimal support for an association of NHL with tobacco exists. A recent study conducted by Morton et al. (2004) included more than 2500 NHL cases from several countries and demonstrated no increase in NHL with tobacco use. A recent pooling of three population-based case control studies in the US also found no association between tobacco and NHL in men, and a weak, non-significant association in women (Zahm et al., 1997).

While several studies detected no association with alcohol, one reported an inverse association between alcohol and NHL (Nelson et al., 1997). Morton et al. (2004) found a significant reduction in NHL risk with alcohol use (RR 0.8, CI 0.7–0.9). This reduced risk was independent of the specific type of alcohol consumed, duration of alcohol use, and lifetime consumption. In a 9-year follow-up study of 35 000 Iowa women, Chiu et al. (1999) also detected a similar inverse association between alcohol intake and NHL risk . Neither alcohol nor tobacco seems to account for the increase in lymphoid malignancies currently being reported.

Blood transfusions

Several studies have demonstrated an association between history of blood transfusion and NHL. Potential biologic mechanisms that may have an etiologic impact include: oncogenic virus transmission, transfusion-induced immunosuppression, and engraftment of malignant lymphoma cells from donor. A population-based study of women in Iowa found a twofold or greater risk of NHL among persons receiving blood transfusions (CI 1.4–3.6) (Cerhan et al., 1993). Three cohort studies have shown a significant increase in risk (RR 1.6–3.5) of NHL with history of blood transfusion, but case control studies have been negative (Blomberg et al., 1993; Memon and Doll, 1994; Cerhan et al., 2001). Latency between transfusion and NHL diagnosis in these studies ranged from less than 6 years to 20 years. Given the likelihood of a multifactorial etiology, the association of transfusions or the underlying condition that necessitated transfusion to NHL risk is not clear. Interestingly, improved renal graft survival has been demonstrated with pre-transplant transfusions (Opelz et al., 1973), suggesting an immunosuppressive effect subsequent to transfusion that is measurable for extended time periods (Landers et al., 1996; Klein, 1999). The immunomodulatory effects of transfusions are proposed to induce a shift in endogenous cytokine secretion and a downregulation of T-cell activity while promoting a Th2 humoral response (Kirkley et al., 1995). It may be appropriate to consider the potential ramifications of mild, subclinical immunocompromised conditions such as that postulated to occur after blood transfusion when combined with an asymptomatic or mild infection on risk of lymphoma.

Summary

NHL rates have risen rapidly in the past three decades. Before advances in the control and prevention of NHL are possible, a better understanding of the origins of the disease is required. The factors that facilitate the initiation and progression of NHL are merely different renditions of the same factors responsible for the regulation of normal lymphocytes. Three common themes underlying the emergence and perpetuation of NHL have been identified: (1) an episodic or persistent immunosuppressive state, which may be the result of primary or acquired immunodeficiency or exposure to an immunosuppressive agent such as UVR or blood transfusion; (2) chronic antigenic stimulation due to an autoimmune condition, viral infection, or allergic/inflammatory agent; and (3) disruption of normal cell proliferation. Within this framework, each neoplasm emanates from a genetic mutation, either a random error or a mutation instigated by an oncogenic agent, and, ultimately, the growth of the malignant cell into an established tumor is promoted by one or more co-factors that interfere with the usual regulatory mechanisms.

Lymphomagenesis may be influenced by genetic, viral, environmental, and biologic factors. While many of these factors have been discussed, it is not certain which factors may serve as an initiator of malignancy and which promote the survival of a cancer cell, perhaps at times merely by ignoring its existence. Regardless, we must build on our current knowledge regarding the etiology of NHL and the mechanisms related to its progression. As knowledge is gained in these two arenas, prevention, treatment, and cure will evolve from promise to reality. Potential strategies that may be feasible in the future include: administration of vaccines and medications (antibiotics or anti-virals) to prevent or treat viral infections that are eventually demonstrated to be associated with NHL; development of immunotherapeutic treatments that will treat existing tumors whose growth may be dependent on the continued presence of a virus, or will enhance the immune response against the tumor; application of therapies that interfere with the factors necessary for tumor survival such as antiangiogenesis drugs and biologics that interrupt critical growth-signaling pathways in the tumor; new approaches to protect or strengthen the immune system against uncontrollable or unidentifiable insults such as vitamins and nutrients that will enhance immune surveillance during periods of increased exposure to carcinogenic factors or of anticipated immunosuppression such as during medical procedures or high stress events; implementation of protective practices for use of substances that are known to be hazardous and mutagenic, that is, protective clothing, air management systems, etc.; and development of new agricultural, industrial, and home products that are effective yet free of known carcinogens.

Interdisciplinary collaborations are needed to investigate the broad scope of factors that may alter individual susceptibility and promote lymphoid malignancies. Cooperation, collaboration, and participation among clinical and laboratory investigators, private and public funding agencies, and, most importantly, individuals diagnosed with NHL, will be critical for the advancement of knowledge and eventual prevention and/or cure of this disease.

References

1. Ambinder RF, Lemas MV, Moore S, Yang J, Fabian D and Krone C. (1999). Cancer Treat. Res., 99, 27–45.
2. American Cancer Society (2001). Cancer Facts and Figures. American Cancer Society: Atlanta.
3. American Cancer Society (2004). Cancer Facts and Figures. American Cancer Society: Atlanta.
4. Baecklund E, Ekbom A, Sparén P, Feltelius N and Klareskog L. (1998). Br. Med. J., 317, 180–181.
5. Banks PM. (1992). Cancer Res., 52 (Suppl), 5453s–54535s.
6. Biggar RJ, Frisch M, Engels EA and Goedert JJ. (2001). J. Acquir. Immune Defic. Syndrome, 26, 371–376.
7. Blair A and Zahm SH. (1991). Med. State Art. Rev., 6, 335–354.
8. Blomberg J, Möller T, Olsson H, Anderson H and Jonsson M. (1993). Eur. J. Cancer, 29A, 2101–2105.
9. Breslin P, Kang HK, Lee Y, Burt V and Shepard BM. (1988). J. Occup. Med., 30, 412–419. | PubMed | ISI | ChemPort |
10. Cerhan JR, Wallace RB, Dick F, Kemp J, Parker AS, Zheng W, Sellers TA and Folsom AR. (2001). Cancer Epidemiol. Biomarkers Prev., 10, 361–368.
11. Cerhan JR, Wallace RB, Folsom AR, Potter JD, Munger RG and Prineas RJ. (1993). Ann. Intern. Med., 119, 8–15.
12. Chiu BCH, Cerhan JR, Gapstur SM, Sellers TA, Zheng W, Lutz CT, Wallace RB and Potter JD. (1999). Br. J. Cancer, 80, 1476–1482. | Article | PubMed | ISI | ChemPort |
13. Cinque P, Brytting M, Wahren B, Linde A, Cinque P, Castagna A, Lazzarin A, Vago L, Parravicini C, Zanchetta N and Monforte AD. (1993). Lancet, 342, 398–401. | Article | PubMed | ISI | ChemPort |
14. Cleary ML and Sklar J. (1984). Lancet, ii, 489–493.
15. Cleghorn FR, Manns A, Falk R, Hartge P, Hanchard B, Jack N, Williams E, Jaffe E, White F and Bartholomew C. (1995). J. Natl. Cancer Inst., 87, 1009–1014. | PubMed | ChemPort |
16. D'Amore F, Hasle H and Hansen KS. (1992). Hematol. Rev., 6, 183–199.
17. David H, Mendoza S, Konishi T and Miller C. (2001). Cancer Lett., 162, 57–64. | Article | PubMed | ISI | ChemPort |
18. Del Maso L and Franceschi S. (2003). Lancet Oncol., 4, 110–116. | Article | PubMed | ISI |
19. De Re V, De Vita S, Marzotto A, Rupolo M, Gloghini A, Pivetta B, Gasparotto D, Carbone A and Boiocchi M. (2000). Blood, 96, 3578–3584. | PubMed | ISI | ChemPort |
20. Devesa SS and Fears T. (1992). Cancer Res., 52 (Suppl), 5432s–54340s. | PubMed | ChemPort |
21. De Vita S, Sacco C, Sanconno D, Gloghini A, Dammacco F, Crovatto M, Santini G, Dolcetti R, Boiocchi M, Carbone A and Zagonel V. (1997). Blood, 90, 776–782.
22. Eltom MA, Jemal A, Mbulaiteye SM, Devesa SS and Biggar RJ. (2002). J. Natl. Cancer Inst., 94, 1204–1210. | PubMed |
23. Ferreri AJ, Guidoboni M, Ponzoni M, De Conciliis C, Dell'Oro S, Fleischhauer K, Caggiari L, Lettini AA, Dal Cin E, Ieri R, Freschi M, Villa E, Boiocchi M and Dolcetti R. (2004). J. Nat. Cancer Inst., 96, 586–594.
24. Filipovich AH, Mathur A, Kamat D and Shapiro RS. (1992). Cancer Res., 52, 5465s–5467s.
25. Fisher MS and Kripke ML. (1982). Science, 216, 1133–1134. | Article | PubMed | ISI | ChemPort |
26. Gaidano G and Dalla-Favera R. (1997). Cancer Principles & Practice of Oncology Vol. 3. DeVita V Jr, Hellman S and Rosenberg SA. (eds). Lippincott-Raven: Philadelphia, NY, pp. 2131–2145.
27. Garabrant DH and Philbert MA. (2002). Crit. Rev. Toxicol., 32, 233–257.
28. Gisbert JP, Garcia-Buey L, Pajares JM and Moreno-Otero R. (2003). Gastroenterology, 125, 1723–1732. | Article | PubMed | ISI |
29. Goedert JJ. (2000). Semin. Oncol., 27, 390–401. | PubMed | ChemPort |
30. Greenlee R, Hill-Harmon M, Murray T and Thun M. (2001). CA: Cancer J. Clin., 51, 15–36. | PubMed | ISI | ChemPort |
31. Grogan T and Miller T. (1995). Cancer Treatment Haskell CM (ed). WB Saunders: Philadelphia, 979–1005.
32. Groves FD, Linet MS, Travis LB and Devesa SS. (2000). J. Natl. Cancer Inst., 92, 1240–1251. | Article | PubMed | ChemPort |
33. Harris NL, Jaffe ES, Diebold J, Flandrin G, Muller-Hermelink HK, Vardiman J, Lister TA and Bloomfield CD. (1999). Ann. Oncol., 10, 1419–1432. | Article | PubMed | ISI | ChemPort |
34. Harris NL, Jaffe ES, Stein H, Banks PM, Chan JK, Cleary ML, Delsol G, De Wolf-Peeters C, Falini B and Gatter KC. (1994). Blood, 84, 1361. | PubMed | ISI | ChemPort |
35. Harris OD, Cooke WT, Thompson H and Waterhouse JAH. (1967). Am. J. Med., 42, 899.
36. Hartge P and Devesa SS. (1992). Cancer Res., 52, 5566s–5569s.
37. Hayes HM, Tarone RE and Cantor KP. (1991). J. Natl. Cancer Inst., 83, 1226–1231.
38. Hoar SK, Blair A and Holmes FF. (1986). JAMA, 256, 1141–1147.
39. Holford TR, Zheng T, Mayne ST and McKay LA. (1992). Cancer Res., 52, 5443s–5446s.
40. Holly EA, Lele C, Bracci PM and McGrath MS. (1999). Am. J. Epidemiol., 150, 375–389. | PubMed | ISI | ChemPort |
41. Hoover RN. (1992). Cancer Res., 52, 5477S.
42. Hussell T, Isaacson P, Crabtree J and Spencer J. (1993). Lancet, 342, 571. | Article | PubMed | ISI | ChemPort |
43. Hussell T, Isaacson PG, Crabtree JE and Spencer J. (1996). J. Pathol., 178, 122–127. | Article | PubMed | ISI | ChemPort |
44. International Agency for Research on Cancer Monographs on the Evaluation of Carcinogenic Risks to Humans (1978). Some Aromatic Amines and Related Nitro Compounds – Hair Dyes, Colouring Agents and Miscellaneous Industrial Chemicals. Monograph 16. International Agency for Research on Cancer: Lyon, France.
45. International Agency for Research on Cancer Working Group on the Evaluation of Carcinogenic Risks to Humans (1996). Human Immunodeficiency Viruses and Human T-cell Lymphotropic Viruses. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans Vol. 67 IARC Press: Lyon, pp. 67.
46. International Agency for Research on Cancer (1997). Epstein–Barr Virus and Kaposi's Sarcoma Herpesvirus/Herpesvirus 8. Monographs on the Evaluation of Carcinogenic Risks to Humans. IARC: Lyon.
47. Jacobson LP, Yamashita TE, Detels R, Margolick JB, Chmiel JS, Kingsley LA, Melnick S and Munoz A. (1999). J. Acquir. Immune Defic. Syndrome, 21 (Suppl 1), S34–S41.
48. Jeevan A and Kripke ML. (1990). Cell Immunol., 130, 32–41. | Article | PubMed | ISI | ChemPort |
49. Jonsson V, Wiik A, Hou-Jensen K, Christiansen M, Ryder LP, Madsen HO, Geisler C, Hansen MM, Thomsen K, Vorstrup S and Svejgaard A. (1999). J. Intern. Med., 245, 277–286.
50. Kamel OW, van de Rijn M, Hanasono MM and Warnke RA. (1995). Leukemia Lymph., 16, 363.
51. Kast WM, Schreiber H and Velders MP. (1998). Encyclopedia of Life Sciences. Macmillan Reference Ltd: UK.
52. Kinlen L. (1992). Cancer Res., 52, 5474s–5476s.
53. Kirkley SA, Cowles J, Pellegrini Jr VD, Harris CM, Boyd AD and Blumberg N. (1995). Lancet, 345, 527. | Article | PubMed | ChemPort |
54. Klein HG. (1999). Anesthesiology, 91, 861–865.
55. Kripke ML. (1994). Cancer Res., 54, 6102–6105. | PubMed | ISI | ChemPort |
56. Landers DF, Hill GE, Wong KC and Fox IJ. (1996). Anesth. Analg., 82, 187–204. | Article | PubMed | ChemPort |
57. Lecuit M, Abachin E, Martin A, Poyart C, Pochart P, Suarez F, Bengoufa D, Feuillard J, Lavergne A, Gordon JI, Berche P, Guillevin L and Lortholary O. (2004). N. Engl. J. Med., 350, 239–248. | Article | PubMed | ISI | ChemPort |
58. Levi F, Randimbison L, Te VC and La Vecchia C. (1996). Br. J. Cancer, 74, 1847–1850. | PubMed | ISI | ChemPort |
59. Levine AM. (1994). Ann. Oncol., 5, 29.
60. Linet MS and Pottern LM. (1992). Cancer Res., 52, 5468s–5473s.
61. Luppi M, Longo G, Ferrari MG, Ferrara L, Marasca R, Barozzi P, Morselli M, Emilia G and Torelli G. (1996). Br. J. Haematol., 94, 373–375. | Article | PubMed |
62. McMichael AJ and Giles GG. (1996). Br. J. Cancer, 73, 945–950. | PubMed | ISI | ChemPort |
63. Melbye M, Adami HO, Hjalgrim H and Glimelius B. (1996). Acta Oncol., 35, 655–657.
64. Melbye M and Trichopoulos D. (2002). Textbook of Cancer Epidemiology Vol. 25. Adami H-O, Hunter D and Trichopoulos D (eds). Oxford University Press: Oxford, pp. 535–555.
65. Memon A and Doll R. (1994). Int. J. Cancer, 58, 366–368.
66. Moore RD and Chiasson RE. (1999). AIDS, 13, 1933–1942. | Article | PubMed | ISI | ChemPort |
67. Morrow RH, Kisuule A, Pike MC and Smith PG. (1976). J. Nat. Cancer Inst., 56, 479–483.
68. Morton LM, Hartge P, Cerhan JR, Holly EA, Holford TR, Mayne ST and Zheng T. (2004). Proc. AACR, 45, 490.
69. Mueller NE, Mohar A and Evans A. (1992). Cancer Res., 52 (Suppl), 5479s–5481s.
70. Murphy EL, Figueroa JP, Gibbs WN, Holding-Cobham M, Cranston B, Malley K, Bodner AJ, Alexander SS and Blattner WA. (1991). Am. J. Epidemiol., 133, 1114–1124. | ChemPort |
71. Nakatsuka S, Lin A, Dong Z, Nomura S, Takakuwa T, Miyazato H, Aozasa K and Osaka Lymphoma Study Group. (2003). Cancer Res., 63, 7606–7608.
72. Nelson RA, Levine AM, Marks G and Berstein L. (1997). Br. J. Cancer, 76, 1532–1537. | PubMed | ChemPort |
73. Offit K, Wong G, Filippa DA, Tao Y and Chaganti RSK. (1991). Blood, 77, 1508. | PubMed | ISI | ChemPort |
74. Opelz G and Henderson R. (1993). Lancet, 342, 1514–1516. | Article | PubMed | ISI | ChemPort |
75. Opelz G, Sengar DP, Mickey MR and Terasaki PI. (1973). Transplant. Proc., 5, 253–259. | PubMed | ISI | ChemPort |
76. Palackdharry CS. (1994). Oncology, 8, 67–73.
77. Palella Jr FJ, Delaney KM, Moorman AC, Loveless MO, Fuhrer J, Satten GA, Aschman DJ and Holmberg SD. (1998). N. Engl. J. Med., 338, 853–860. | Article | PubMed | ISI |
78. Paltiel O, Schmit T, Adler B, Rachmilevitz EA, Polliack A, Cohen A, Haim N, Shachar MB, Epelbaum R, Barchana M, Cohen R and Ben Yehuda D. (2000). Cancer, 88, 2357–2366. | Article | PubMed | ISI | ChemPort |
79. Parsonnet J, Hansen S, Rodriguez L, Gelb AB, Warnke RA, Jellum E, Orentreich N, Vogelman JH and Friedman GD. (1994). N. Engl. J. Med., 330, 1267–1271. | Article | PubMed | ISI | ChemPort |
80. Pearce N. (1989). Br. J. Ind. Med., 46, 143–144.
81. Pettersson T, Pukkala E, Teppo L and Friman C. (1992). Ann. Rheum. Dis., 51, 437–439.
82. Pimental D, McLaughlin L and Zepp A. (1993). Handbook of Pest Management in Agriculture Pimental D (ed), Vol. 1 CRC Press: Boca Rotan, pp. 679–680.
83. Rabkin CS, Biggar RJ and Horm JW. (1991). Int. J. Cancer, 47, 692–696.
84. Reed DM. (1902). Johns Hopkins Hosp. Rev., 10, 133.
85. Ries L, Eisner M, Kosary C, Hankey B, Miller B, Clegg L and Edwards B (eds) (2001). SEER Cancer Statistics Review. National Cancer Institute: Bethesda, MD, 1973–1998.
86. Rivas JM and Ullrich SE. (1992). J. Immunol., 149, 3865–3871. | PubMed | ISI | ChemPort |
87. Rizzo P, Carbone M, Fisher SG, Matker C, Swinnen LJ, Powers A, Di Resta I, Alkan S, Pass HI and Fisher RI. (1999). Chest, 166 (Suppl), 470S–473S.
88. SEER (2003). Surveillance, Epidemiology & End Results Program Public-Use Data 1973–2000, National Cancer Institute, DCCPS, Surveillance Research Program, Cancer Statistics Branch (released May).
89. Selik RM, Starcher ET and Curran JW. (1987). AIDS, 1, 175–182.
90. Shipp MA, Mauch PM and Harris NL. (1997). Cancer Principles & Practice of Oncology Vol. 3. DeVita Jr V, Hellman S and Rosenberg SA (eds). Lippincott-Raven: Philadelphia, New York, pp. 2165–2220.
91. Shivapurkar N, Harad K, Reddy J, Scheuermann RH, Xu Y, McDenna RW, Michgrub S, Kroft SH, Feng Z and Gasdar AF. (2002). Lancet, 359, 851–852. | Article | PubMed | ISI | ChemPort |
92. Swinnen L, Costanzo-Nordin MR, Fisher SG, O'Sullivan EJ, Johnson MR, Heroux AL, Dizikes GJ, Pifarre R and Fisher RI. (1990). N. Engl. J. Med., 323, 1723. | PubMed | ISI | ChemPort |
93. Takatsuki K, Uchiyama T, Sagawa K and Yodoi J. (1976). Topics in Hematology, Adult T Cell Leukemia in Japan. Seno S, Takaku F, Iron S (eds). Excerpta Medica: Amsterdam, pp. 73–78.
94. Tavani A, La Vecchia C, Franceschi S, Serraino D and Carbone A. (2000). Eur. J. Cancer Prev., 9, 59–64. | Article | PubMed | ISI | ChemPort |
95. Tycko B and Sklar J. (1990). Cancer Cells, 2, 1. | PubMed | ISI | ChemPort |
96. Vermeer M and Streilein JW. (1990). Photodermatol. Photoimmunol./Photomed., 7, 258–265.
97. Vilchez RA, Madden CR, Kozinetz CA, Halvorson SJ, White ZS, Jorgensen JL, Finch CJ and Butel JS. (2002). Lancet, 359, 817–823. | Article | PubMed | ISI | ChemPort |
98. Vink AA, Yarosh DB and Kripke ML. (1996). Photochem. Photobiol., 63, 383–386. | PubMed | ISI | ChemPort |
99. Wahl SM, Wilder RL, Katona IM, Wahl LM, Allen JB, Scher I and Decker JL. (1983). Arthritis Rheum., 26, 1076–1084.
100. Wiernik PH, Wang SQ, Hu XP, Marino P and Paietta E. (2000). Br. J. Haematol., 108, 72–79.
101. Wigle DT, Semenciw RM, Wilkins K, Riedel D, Ritter L, Morrison HI and Mao Y. (1990). J. Natl. Cancer Inst., 82, 575–582. | PubMed |
102. Willemze R, Kerl H, Sterry W, Berti E, Cerroni L, Chimenti S, Diaz-Peréz JL, Geerts ML, Goos M, Knobler R, Ralfkiaer E, Santucci M, Smith N, Wechsler J, van Vloten WA and Meijer CJLM. (1997). Blood, 90, 354–371. | PubMed | ISI | ChemPort |
103. Wotherspoon AC, Ortiz-Hidalgo C, Falzon MR and Isaacson PG. (1991). Lancet, 338, 1175–1176. | Article | PubMed | ISI | ChemPort |
104. Ye BH. (2002). Cancer Investig., 18, 356–365.
105. Yosida M, Kiyokawa T, Watanabe T, Hattori S, Fujisawa J-I and Seiki M. (1985). Retroviruses in Human Lymphoma/Leukemia. Miwa M (ed). Scientific Societies Press: Tokyo, Japan, pp. 39–47.
106. Zahm SH, Weisenburger DD and Babbitt PA. (1990). Epidemiology, 1, 349–356. | PubMed |
107. Zahm SH, Weisenburger DD, Holmes FF, Cantor KP and Blair A. (1997). Cancer Causes Control, 8, 159–166. | Article | PubMed | ChemPort |
108. Zhang Y, Holford TR, Leaderer B, Boyle P, Zahm SH, Flynn S, Tallini G, Owens PH and Zheng T. (2004). Am. J. Epidemiol., 159, 148–154.
109. Zheng T, Mayne ST, Boyle P, Holford Tr, Liu WL and Flannery J. (1992). Cancer, 70, 840–849. | PubMed |
110. Zhu K, Levine RS, Gu Y, Brann EA, Hall I, Caplan LS and Baum MK. (1998). Cancer Causes Control, 9, 77–82.