Epstein-Barr Virus as a Cause of Human Disease
Worldwide infection with EBV is so widespread that by the third decade of life it is estimated that 80% to 100% of individuals have become carriers of the infection (IARC, 1997). In the United States, antibody to EBV occurs in about 10% of individuals in the neonatal period, rising to some 60% by 10 years of age (Evans and Niederman, 1989). In contrast, antibody to EBV is seen in nearly 90% to 100% of individuals less than 10 years of age in Africa as well as the East and West Indies. In parallel with these differences, the picture of EBV-induced disease differs significantly in the United States and these other regions of the world. Perhaps the most common disease induced by EBV in the United States is infectious mononucleosis, characterized by fe- ver, sore throat, tonsillitis, lymphadenopathy, and splenomegaly. This self-limiting disease rarely has a protracted course, although some complications may occur (Sumaya, 1989; Okano and Gross, 1996). Primary EBV infection is usually asymptomatic, and infection occurs predomi- nantly via salivary and urinary contact (cf. Gratama and Ernberg, 1995). Thus, in populations with high standards of hygiene, such as predominate in western cultures, primary EBV infec- tions may not occur until late childhood or adolescence, at which time the potential for develop- ing infectious mononucleosis appears to predominate in individuals with a relatively mature immune system. In contrast, in Third World countries, especially where classic Burkitt lym- phoma predominates, extensive infections with the virus occur early in life, as noted by the anti- body response, resulting in neoplastic development. This theory, while not yet proven, does take into account a number of the epidemiological and geographical characteristics of these diseases.
The clinical and histopathological characteristics of Burkitt lymphoma have allowed the distinction of at least two forms of the disease—that predominating in young children and af- fecting lymphoid tissue of the head and neck, with quite characteristic clinical features in associ- ation with the lymphadenopathy of the disease; and a second clinical form involving lymphoid structures in the abdominal cavity, seen in older children and adults (cf. Levin and Levine,1998). In the first or “African” form of the disease, more than 95% of patients exhibit infection with the Epstein-Barr virus and infectious virus within the neoplasm. In the abdominal form, which is increasing in Africa along with socioeconomic improvements, only 10% to 30% of such neoplasms appear to be associated with the Epstein-Barr virus (IARC, 1997). There is also substantial evidence that malarial infestation, especially in the young, increases the risk of devel- oping Burkitt lymphoma. The lymphoma is rare in regions in which malaria eradication cam- paigns have been successful. Another possible environmental factor in the pathogenesis of Burkitt lymphoma in Africa is the presence of and use of products from the plant Euphorbia tirucalli (IARC, 1997). The phorbol esters present in this plant have structures closely related to the phorbol esters used as tumor-promoting agents in experimental studies in mouse epidermal carcinogenesis (Chapter 7). Such agents can increase the ability of EBV to transform B lympho- cytes and increase the likelihood of chromosomal alterations in such cells. The plant is endemic in areas of equatorial Africa where Burkitt lymphoma incidence is quite high.
EBV is also causally associated with the development of nasopharyngeal carcinoma, a dis- ease with a remarkable racial and geographical distribution. In most parts of the world, the inci- dence is usually less than 1 per 100,000 persons per year (Parkin et al., 1997). However, in some provinces of mainland China, the rates may be as high as 25 to 40 per 100,000 person-years. The most common histological type of these neoplasms is undifferentiated carcinomas, with promi- nent lymphoid stroma accounting for up to 80% of the cases (Niedobitek et al., 1996). Squamous cell carcinomas with little or no lymphoid stroma make up the remaining percentage of these neo- plasms. While the undifferentiated carcinomas with lymphoid stroma almost all exhibit detectable virus, a smaller percentage of the squamous cell carcinomas do so. Thus, even sporadic cases throughout the world of the undifferentiated carcinoma/lymphoid stroma type of nasopharyngeal carcinoma exhibit molecular evidence of EBV causation. To a greater extent than with Burkitt lymphoma, environmental factors appear to regulate the expression of the neoplastic state in that individuals living in endemic regions migrating to nonendemic regions exhibit a dramatic de- crease in the incidence of nasopharyngeal carcinoma within the first generation (IARC, 1997). Several environmental factors, including the consumption of highly salted fish and other preserved foods, have been related to nasopharyngeal carcinoma incidence, specifically in the regions where this lesion is endemic. In addition, exposure to formaldehyde, tobacco smoke, alcohol, and vari- ous herbal preparations as well as host genetic factors have been implicated in this dramatic mi- gratory effect on the incidence of nasopharyngeal carcinoma (Jeannel et al., 1999).
Hodgkin disease is a form of malignant lymphoma characterized by the presence of abnor- mal mononuclear cells and their multinucleated variants, termed Reed-Sternberg cells. In addi- tion, the histological picture is quite heterogeneous and variable, with various classifications having been reported (Jackson and Parker, 1944; Lukes et al., 1966). The incidence of Hodgkin disease in western populations is largely bimodal with respect to age (Figure 12.9). Generally, few cases occur among children, but there is a rapid increase in incidence among teenagers, peaking at about age 25. The incidence then decreases to a plateau through middle age, after which rates increase with age to a second peak. In developing countries, Hodgkin disease in childhood is much more common, with the children at risk appearing to be of the lower social class. Epidemiological studies have suggested a role for the Epstein-Barr virus in the causation of Hodgkin disease, since infectious mononucleosis was found to be a significant risk factor for Hodgkin disease in the first 5 years after diagnosis of the former condition (cf. Levin and Le- vine, 1998). Furthermore, infectious mononucleosis usually occurs in young adulthood, in higher social classes with higher maternal education and lower housing density in childhood; in addition, smaller sibship size is another risk factor for young adulthood Hodgkin disease, all of these factors being indicative of a “late” exposure to an infectious agent (Mueller, 1996). Al- though a relationship between EBV and Hodgkin disease had been suspected for several de- cades, the breakthrough for such a causative association came from molecular epidemiology. Assays by a variety of techniques have demonstrated the EBV genome in neoplastic cells of about 30% to 50% of Hodgkin disease cases in western countries; but in South America and the Orient, this proportion may reach 60% to 100%. Thus, the infectious pattern for this disease is quite analogous to that for Burkitt lymphoma in the younger age groups, although the pathogen- esis of the EBV induction of Hodgkin disease in older age groups is not clear.
Another relatively unusual neoplasm strongly associated with EBV infection is the sino- nasal T-cell lymphoma, also known as lethal midline granuloma or reticulosis. EBV is present in most tumor cells of these neoplasms (IARC, 1997) but is found much less frequently in sinona- sal B-cell lymphomas. This neoplasm indicates that EBV may colonize T cells as well as B cells, the latter characteristic of infectious mononucleosis and Burkitt lymphoma. Several other neoplasms have also been associated with EBV with molecular technologies, but for the most part at very low incidence. These include gastric adenocarcinoma and a variety of lymphoepithe-
Figure 12.9 Age-specific incidence rates of Hodgkin disease per 100,000 population for each sex as noted in the areas listed in each of the four graphs. The inclusive years studied were as follows: Cali-Co- lombia, 1962–66; Connecticut, U.S.A., 1960–62; Rural Norway, 1964–66; Urban Norway, 1964–66. (Adapted from IARC, 1997, with permission of the authors and publisher.)
liomas seen in the human (Anagnostopoulos and Hummel, 1996). In addition, EBV may infect certain New World primates, particularly cotton-topped tamarins, with the resultant develop- ment of lymphomas (IARC, 1997).