CHAPTER 10 Epstein-Barr Virus Alfred S. Evans and James C. Niederman 1. Introduction 2. Historical Background Epstein-Barr virus (EBV), a member of the herpes In 1889, Emil Pfeiffer of Wiesbaden, Germany, group of viruses, is the cause of heterophile-positive described a condition called Driisenfieber (glandular infectious mononucleosis, of most heterophile-neg­ fever), characterized by fever, adenopathy, mild ative cases, and of occasional cases of tonsillitis and sore throat, and in severe cases enlargement of the pharyngitis in childhood. Rarely, it may involve the liver and spleen. (63, 126) Since Pfeiffer's original de­ liver or central nervous system as primary mani­ scription, as well as Filatov's, (53,54) a Russian, in festations. This virus is strongly implicated as hav­ 1892, antedated by some 30-50 years recognition of ing a causal relationship to African Burkitt lym­ the hematological changes and the heterophile an­ phoma and to nasopharyngeal cancer. High antibody tibody, it is uncertain whether these were true in­ titers are also present in 30-40% of cases of Hodg­ fectious mononucleosis. However, his description kin's disease, in some patients with sarcoidosis, and of this febrile syndrome in older children and young in systemic lupus erythematosus. adults seems best to fit this diagnosis. There is little This chapter will deal with the epidemiology of doubt about the classic description of the disease EBV infections and the epidemiology of infectious made by Sprunt and Evans, (139) from Johns Hop­ mononucleosis. Infectious mononucleosis can be kins, in 1920. They described the disorder in young defined as an acute febrile illness involving children adults as we now know it, named the disease "in­ and young adults characterized clinically by sore fectious mononucleosis," and reported in detail the throat and lymphadenopathy, hematologically by hematological changes. This description was fol­ lymphocytosis of 50% or more, of which 10% or lowed rapidly by similar reports from other work­ more are atypical, and serologically by an elevated ers.(12,14,22,96,108) A definitive presentation of hema- absorbed heterophile-antibody titer and the devel­ tological changes was made by Downey and opment of EBV immunoglobulin M (IgM) and other McKinlay(28) in 1923. The next major development EBV antibodies. This chapter also mentions the re­ was the discovery in 1932 of the heterophile anti­ lationship of high antibody titers to certain chronic body by John R. Paul and William W. Bunnell(121) and malignant diseases, but the major discussion of Yale University. Their report was based on an of Burkitt lymphoma and of nasopharyngeal cancer accidental observation while studying the occur­ will be found in later chapters of this book. rence of heterophile antibodies in rheumatic fever. Alfred S. Evans . WHO Serum Reference Bank, Section This search had been initiated because of the clinical of International Epidemiology, Department of Epidemiol­ similarity of rheumatic fever and serum sickness and ogy and Public Health, Yale University School of Medicine, because of the work of Davidsohn(24) describing the NewHaven,Connecticut. JamesC.Niederman . De­ partment of Epidemiology and Public Health, Yale Uni­ presence of heterophile antibodies in serum sick­ versity School of Medicine, New Haven, Connecticut. ness. Among the control subjects for rheumatic 253 A. S. Evans (ed.), Viral Infections of Humans © Plenum Publishing Corporation 1982 254 Chapter 10 • Epstein-Barr Virus fever patients was one who had infectious mono­ fectious agent from the throat or blood of patients nucleosis and was found to have a much higher with infectious mononucleosis using several tissue­ heterophile antibody titer than present in any other culture systems, long-term cultures of lymphocytes condition. Paul and Bunnell then continued these on a feeder layer, and fluorescent antibody tech­ observations in 3 additional cases of infectious niques to identify an agent were unsuccessful. (37) mononucleosis and utilized 275 controls for com­ EpidemiolOgically, the key events during this time parison. Their paper also describes what they be­ were the observations of Hoagland, who suggested lieved to be a false-positive heterophile antibody that the disease might be transmitted by kissing(79) occurring in a patient with aplastic leukemia. A re­ and that the incubation period was of the order of view of the details of this case(43) reveals that the 30-49 days. (81) heterophile antibody occurred about 20 days after Early in 1968, evidence first appeared that EBV the administration of several units of blood, and was the cause of infectious mononucleosis. (72,114) therefore this patient may represent the first case This virus, isolated by Epstein et al. (33) from a culture of transfusion infectious mononucleosis. Soon after of African Burkitt tumor tissue, (128) was found to be the discovery of the presence of heterophile anti­ a new member of the herpes group of viruses. While bodies, Davidsohn and Walker<23) reported on the working with this agent, a technician in the Henles' use of guinea pig kidney and of beef cells to absorb laboratory in Philadelphia developed infectious serum prior to heterophile testing in order to in­ mononucleosis. Her serum, which lacked antibody crease the specificity of the test. Both these proce­ several months prior to disease, developed EBV an­ dures have withstood the test of time well and still tibody during illness, and her lymphocytes, which constitute one of the major criteria of diagnosis. had previously failed to be cultivated successfully, Regular alterations in various liver-function tests now grew well in tissue culture and were shown to during acute infectious mononucleosis were rec­ contain EBV antigen. (72) This serendipitous obser­ ognized in several laboratories in the late 1940s and vation was rapidly confirmed and extended at that 1950s, (13,35,86) even though only 5% of patients had time by the Henles in conjunction with Niederman clinical jaundice. This was followed by the discovery and McCollum(114) of Yale and later in several other of alterations in serum glutamic oxalacetic trans­ prospective studies carried out by the Yale aminase (SCOT) and other hepatic enzymes during team(SO,113,132) and in one English study. (148) Subse- the course of disease. (6,131,158) quent investigations established the presence and Search for the etiological agent of infectious mono­ persistence of EBV in the throat during and after nucleosis began in the 1920s, but met with little acute infectious mononucleosis,<t9,59,103) the occur­ success until 1942, when Wising(156) reported the rence of EBV-specific IgM, (3,66,114b,133) and the re­ successful transmission of classic infectious mono­ production of some features of mononucleosis by nucleosis to a female medical student volunteer inoculation of EBV into monkeys.(135,154) Fuller de­ who received 250 ml of blood from a patient ill with tails of the history of infectious mononucleosis have the acute disease. This successful experiment was been published elsewhere. (16,43) not reproducible by Wising in several other at­ tempts, nor by Bang, (4) who carried out a similar set of volunteer experiments. In 1947 and again in 1950, additional efforts of this sort were carried out at Yale 3. Methodology University using whole blood, serum, or throat washings. The results provided suggestive but in­ 3.1. Mortality Data conclusive evidence of transmission. (34,36) A third effort without success was reported from Yale Uni­ Infectious mononucleosis is rarely a fatal disease; versity in 1965. (112) Subsequent EBV antibody tests only about 50 fatalities have been reported. (39) Ex­ on the sera from these last experiments in 1968 re­ amination of autopsy records or of international in­ vealed that all volunteers had actually been immune dexes of causes of death would therefore give little to infection prior to the experiment as indicated by indication of the occurrence of the disease even the presence of antibody. (111) though its pathological features are quite character­ Repeated attempts in the 1950s to isolate an in- istic. Chapter 10 • Epstein-Barr Virus 255 3.2. Morbidity Data way, an elevated titer quite accurately reflects the occurrence of infectious mononucleosis even in the Infectious mononucleosis is not a reportable dis­ absence of clinical and hematological data and has ease in most states and in most countries. Excep­ been utilized as an indicator of infectious mononu­ tions are the state of Connecticut, where it has been cleosis in sera sent to hospitals and state labora­ reportable since 1948}21) and the United States tories. (38) The major limitation of this approach is armed forces, which collect hospitalization data on the extent to which physicians have sent sera from all diseases. (41) Unless strong emphasis is placed on suspected cases to the diagnostic laboratory for the need for fulfilling clinical, hematological, and analysis. The increasing use of simple laboratory serological criteria for diagnosis before reporting, kits for identifying heterophile-antibody elevations the reliability of morbidity data from these sources in the physician's office probably results in much must be seriously questioned. This requirement is less utilization of state and hospital laboratories, so emphasized by the fact that even for the 15-25 age that morbidity data from these sources may greatly group-in which the disease has its highest inci­ underestimate the occurrence of the disease.