(w; A

n n a ls of

C

l in ic a l a n d

................ L

aboratory

S c ie n c e ,

Copyright © 1974, Institute for Clinical Science

Vol. 4, No. 2

The Epstein-Barr Virus (EBV) in B urkitt’s Lym phom a and N asopharyngeal Carcinom a* W E R N E R H E N L E , M .D .f A N D G E R T R U D E H E N L E

Division of Virology, The Childrens Hospital of Philadelphia, Philadelphia, PA 19146 and School of Medicine, University of Pennsylvania Philadelphia, PA 19104

ABSTRACT Evidence is presented for an oncogenic potential of Epstein-Barr virus (E B V ). Information is also given on how EBV-specific test procedures might aid in the diagnosis of Burkitt’s lymphoma and nasopharyngeal carcinoma and might provide information on the prognosis of the patients and the success of therapy.

One might question whether or not this presentation should be included in a sem­ inar on the laboratory diagnosis of cancer. The answer depends on whether or not one accepts the mounting evidence for an etiologic involvement of the Epstein-Barr virus (E B V ) in Burkitt’s lymphoma (B L ) , the most frequent malignancy of African chil­ dren, and in nasopharyngeal carcinoma (N P C ) of adults in many parts of the world. EBV, a member of the herpes group of viruses, has a world-wide distribution based on antibody surveys. Primary EBV infections at an early age either remain silent or usually cause such mild illnesses that they are not readily distinguishable

from many other minor viral diseases of childhood. With advancing age, primary EBV infections increasingly induce typical infectious mononucleosis (IM ) and, in­ deed, the virus was shown to be the cause of this disease. The evidence for this rela­ tionship will not be discussed here since it has been reviewed recently.5’6 IM is a lymphoproliferative disease which has earned the epithet of a self­ limited leukemia since, in its early stages, it may be confused with such a malignancy. It thus is not inconceivable that EBV-stimulated lymphoid cells might escape, on rare occasions, control by host defenses and cause true malignancies. In fact, EBV is, as a rule, not completely eliminated after the primary infection but establishes a per­ manent carrier state in the lymphoreticular system. Under low socio-economic conditions, such as prevailing in African regions of high B L incidence, primary EBV infections

* Supported by Research Grant CA 04568 and Contract PH 43-66-477 within The Virus Cancer Program, National Cancer Institute, U. S. Public Health Service. f Career Award 5-K6-22, 683, National Insti­ tutes of Health, U. S. Public Health Service. 109

110

H E N L E AN D H E N L E

In view of these results, it is not surpris­ ing that all continuous cultures initiated with biopsy cells invariably contain EBV. Such lines are presently indistinguishable from continuous lymphoblast cultures es­ tablished from peripheral leukocytes of IM patients or healthy carriers of the virus. They are divided into “producer” and “non­ producer cultures” depending on the con­ tinuing presence or absence of small per­ centages of lymphoblasts which synthesize virus particles, or at least viral capsid anti­ gens (V CA ) and/or early viral antigens (E A ). All cells, whether or not from pro­ ducer or non-producer lines and regardless of their origins, were shown to contain EBN A which thus appears to be the earliest expression of the viral genomes they carry. Injection of cells from such lines into immunoincompetent or immunosuppressed animals was found to cause fatal, metastasizing tumors. (2 ) EBV, whether or not derived from producer cultures of BL, IM or other or­ igins, was shown to transform lymphoid cells from susceptible donors, cord blood or fetal organs in vitro into permanently growing lymphoblasts. Lines so established have the same morphologic appearance and (1 ) Fingerprints of EBV were found in range of properties as the continuous cul­ nearly every biopsy examined; that is (a ) tures initiated with B L cells or IM leuko­ viral DNA was demonstrated in amounts cytes discussed previously. These findings equivalent to multiple EBV genomes per indicate that EBV has an oncogenic po­ cell by hybridization of cellular DNA with tential. 3H-labelled viral DNA or c-RNA tran­ (3 ) Early attempts to induce tumors in scribed therefrom,14’20’21 (b ) EBV-deternon-human primates and other animals mined cell membrane antigens (M A ) were by injection of EBV-containing materials detected by immunofluorescence in the failed owing, among other reasons, to spe­ majority of live biopsy cells unless the cells cies barriers or specific immunity since some primate species were found to ac­ were coated by the patient’s antibodies quire antibodies to EBV under natural con­ blocking the antigenic sites12 and (c ) an ditions. However, fatal lymphoproliferative EBV-associated nuclear antigen (E B N A ), malignancies were induced recently in sev­ comparable to the T antigens in cells trans­ eral marmosets within five to six weeks formed by papovaviruses or adenoviruses after injection of EBV derived originally was demonstrated by an anticomplement from an IM patient18 and in one owl mon­ immunofluorescence technique in practi­ key within 14 weeks after injection of cul­ cally every cell of given biopsies.16

occur with few exceptions within the first three years of life. Under improved socio­ economic conditions, primary EBV infec­ tions are increasingly delayed to older age groups; ultimately, however, nearly every­ body becomes infected and, with it, a per­ manent carrier of the virus. These observa­ tions suggest that B L might develop, under rare circumstances, as a relatively early consequence of primary EBV infections, a possibility now explored in a prospective serologic study of the disease in East Africa. In contrast, NPC, being mainly a tumor of adults, would have to be a late, rare consequence of persistent viral carrier states. If these considerations were valid, other factors undoubtedly must contribute to induction of these malignancies by EBV. These factors could be immunologic de­ fects, genetic predisposition, environmental co-carcinogens, or preconditioning by other viral or parasitic infections. The evidence for an etiologic association of EBV with B L has been reviewed re­ cently7’11 and, therefore, only the most per­ tinent and recent references will be cited. The evidence can be divided into four categories:

E B V IN B U B K IT t ’ s L Y M P H O M A AN D N A SO P H A B Y N G E A L C A R C IN O M A

111

TABLE I E B V - I I e l a t e d A n t ig e n - A n t ib o d y S y s t e m s

Antibodies Versus

Test Cells from Producer (P) or Non-producer (NP) Lines

VCA - Viral capsid antigens MA - Cell membrane antigens

P.- Acetone-fixed smears P - Live

EA

NP- Abortively EBV-infected Acetone- or methanolfixed smears Acetone-fixed smears N P , extracts NP- Acetone-methanolfixed smears NP- Live

D

- Early antigens - Diffuse component

R - Restricted component S - Soluble antigens EBNA- Nuclear antigen Infectious virus particles

Test Procedure

Indirect IF * 3 Blocking of direct IF 2 Indirect IF1*

C' fixation 15 Anti-C* IF 16 Neutralization (no inhibition of' colony formation 17

♦Immunofluorescence

tured B L cells from a producer line.|: These findings, if confirmed and shown to be a direct and specific effect of EBV, offer the prospect of animal models for study of the oncogenicity of EBV. (4 ) All African B L patients were found to have antibodies to EBV as did most of appropriate control children. However, striking differences between the two groups became evident by comparing titers and the spectra of antibodies to the various EBV-related antigens listed in table I. Every patient had anti-VC A, often at high titers, so that the geometric mean was eight-fold higher than that of controls of which a few had no anti-VCA. Similar quantitative differences were noted with respect to anti-MA and EBV-neutralizing antibodies. Even more striking is the fact that control children rarely had anti-EA and, if so, mostly anti-R at very low titers. In contrast, many B L patients had high | Epstein, M. A., personal communication,

anti-R titers. Some had additional or only anti-D titers. Anti-R cannot be detected in the presence of brilliant staining by anti-D. In longitudinal serologic studies of B L patients, changes in the spectra and titers of antibodies were found referable to pre­ ceding or subsequent changes in clinical status and, in turn, to provide prognostic information. Loss of anti-MA preceded in several patients by several months recog­ nition of recurrent tumors. Patients in re­ mission who had no anti-EA or who showed steadily declining anti-R titers were those who became long-term survivors.8 Over 35 patients who have survived, to date, 5 to 10 years without evidence of disease showed these patterns. In contrast, patients in remission who maintained, or in time developed high anti-R and/or anti-D titers had usually multiple, ulti­ mately fatal tumor recurrences. The great majority of patients who died six or more months after initial admission had high

1 12

H E N L E AN D H E N L E

anti-R and often also anti-D titers at the time of death. Since D and R are intra­ cellular antigens which, furthermore, are not expressed detectably in the tumor, the corresponding antibodies serve merely as indicators of other adverse conditions such as a simultaneous presence or development of tumor enhancing antibodies or declines or loss of cell mediated immunity. Evidence for an etiologic association of EB V with NPC also has been steadily in­ creasing.7’11 It can be summarized thus: (1 ) Biopsies were shown to contain reg­ ularly viral DNA at concentrations similar to those found in B L specimens.14’20’21 While the viral genomes could be restricted to the lymphoid elements of the tumor, recent studies on frozen tumor sections in­ dicated that they might be present mainly in the carcinoma cells:19 (a ) in situ hy­ bridization experiments with 3H-labelled viral c-RNA followed by autoradiography showed that the isotope was concentrated in what appeared to be carcinoma cells and (b ) tests for the presence of EBN A yielded suggestive staining. Unfortunately, the car­ cinoma cells do not survive in culture to permit confirmation of these results under less complex conditions. (2 ) Since no cells other than those of the lymphoid series have been transformed by EBV, it is unlikely that NPC is a direct consequence of primary infections as sug­ gested also by the seroepidemiologic data. However, viable hybrid cells were obtained in vitro by para-influenza virus-induced fu­ sion of epitheloid cells with lymphoblasts from producer lines.1 The hybrid cells re­ tained E B viral genomes during prolonged cultivation as evident from demonstration of viral DNA sequences in cellular DNA extracts and EA synthesis in some of the cells following derepression of the genomes by exposure of the cultures to 5-iododeoxuridine (Id U ) or 5-bromodeoxuridine (B rd U ). It is tempting to speculate that such hybrid cells might arise during para­

influenza or other lytic virus infections in the postnasal space of EB V carriers by fusion of epithelial cells with EB V genomecarrying lymphoblasts that might be pres­ ent at the site at the right time. This hy­ pothesis might be testable in non-human primates. (3 ) All NPC patients were found to have antibodies to EBV , but the titers and spectra of antibodies were related to the stage of the disease; i.e., the total tumor burden.9 In Stage I, in which no lymph node involvement is noted, the anti-VC A titers were at most slightly higher than in healthy control populations and anti-EA was usually absent. With progress of the disease to Stages II to V accompanied by lymph node invasion at increasingly more distant sites from the primary tumor and ultimately wide-spread metastases, the geo­ metric mean anti-VCA titers increased stepwise to a finally 8-fold higher level than observed in Stage I patients. Also anti-EA; i.e., dominantly anti-D, became detectable in Stage II and reached high titers in later stages. Conversely, after suc­ cessful eradication of the tumor, the antiVCA titers declined to lower levels and anti-D often became non-detectable again. These results denote that progression of the disease is reflected by rising antibody titers and that the success of therapy can be monitored serologically, especially by ob­ servation of the anti-D reactivity. The experience with the B L and NPC patients has shown that longitudinal sero­ logic studies may provide information on the progression of the diseases, the likeli­ hood of relapses and the effectiveness of therapy. Tests for EBV-related antibodies might therefore be of some value in the therapeutic management of patients. From a diagnostic point of view, EBV-specific test procedures hardly seem needed since the diagnosis of these two malignancies is established on clinical and histological grounds. However, problems have arisen

E B V I N B U B K IT T S L Y M P H O M A AND N A SO P H A R Y N G EA L C A RCIN O M A

with respect to B L . Biopsies of two recent African patients failed to reveal EBV-DNA or EBNA.§ Cells from these biopsies con­ tained, however, a nuclear antigen detecta­ ble by the anti-complement immunofluores­ cence technique with sera from patients with acute myelogenous leukemia. If this antigen were, like EBNA, the fingerprint of a virus, it would appear that B L could be induced occasionally by a virus other than EBV. This could explain why a con­ siderable proportion of American or Euro­ pean patients with a diagnosis of B L had no antibodies to EBV, or anti-VCA titers not exceeding those generally found in con­ trols.10’13 The diagnosis of some of the C au­ casian patients studied in this laboratory was probably incorrect since the tumor arose at sites not observed in African cases, or did not respond significantly to therapy and, in fact, was reclassified in several in­ stances at autopsy as another malignancy. Other Caucasian patients conformed, how­ ever, in every aspect to African cases, in­ cluding EBV-related serology. Search for EBN A and other as yet unidentified or un­ known nuclear antigens might serve in the future to establish probable causes of this type of tumor. Also NPC in Caucasians appears to differ to some degree from Chinese or E ast Af­ rican cases. 11 While all patients have shown high anti-VCA titers, some American or Swedish patients showed low or no anti-D titers. This could denote that in Caucasians the tumor is generally recognized at an early stage or that it is frequently of the invasive type in which lymph node involve­ ment is limited or absent. Chinese patients with the invasive type of NPC had low or no detectable levels of anti-D even when classified as having Stage III or IV disease.9 The differences observed would limit the § Klein, G., personal communication. || Henderson et al, de Schryver et al, to be published.

113

usefulness of EBV-specific serology in Cau­ casian NPC patients. R eferen ces 1. G l a s e r , R ., N o n o y a m a , M., D e c k e r , B., a n d R a p p , F .: Synthesis of Epstein-Barr virus anti­ gens and DNA in activated Burkitt somatic cell hybrids. Virology, in press. 2. G u n v é n , P. a n d K l e i n , G . : Blocking of direct membrane immunofluorescence in titration of membrane-reactive antibodies associated with EBV . J. Nat. Cancer Inst. 47:539-548, 1971. 3. H e n l e , G. a n d H e n l e , W. : Immunofluores­ cence in cells derived from Burkitt’s lym­ phoma. J . Bacteriol. 9 1 :1248-1256, 1966. 4. H e n l e , G., H e n l e , W., a n d K l e i n , G .: Dem­ onstration of two distinct components in the early antigen complex of Epstein-Barr virusinfected cells. Int. J . Cancer 8:272-278, 1971. 5. H e n l e , W. a n d H e n l e , G .: Epstein-Barr virus: The cause of infectious mononucleosis— A review. Oncogenesis and Herpesviruses. Biggs, P. M., de-Thé, G., and Payne, L . N., eds., Int. Agency Res. Cancer, pp. 269-274, 1972. 6 . H e n l e , W. a n d H e n l e , G. : Epstein-Barr virus and infectious mononucleosis. Editorial. New Eng. I. Med. 288:263-264, 1973. 7. H e n l e , W. a n d H e n l e , G .: Evidence for an oncogenic potential of the Epstein-Barr virus. Cancer Res. 33:1419-1423, 1973. 8 . H e n l e , W., H e n l e , G ., G u n v é n , P., K l e i n , G ., C l i f f o r d , P., a n d S i n g h , S . : Patterns of antibodies to Epstein-Barr virus-induced early antigens in Burkitt’s lymphoma. Comparison of dying patients with long-term survivors. J. Nat. Cancer Inst. 50:1163-1173, 1973. 9. H e n l e , W., H o , H - C ., H e n l e , G ., a n d K w a n , H. C .: Antibodies to Epstein-Barr virus-re­ lated antigens in nasopharyngeal carcinoma. Comparison of active cases with long-term survivors. J . Nat. Cancer Inst. 51:361-369, 1973. 10. H i r s h a u t , Y., C o h e n , M. H ., a n d S t e v e n s , D. A.: Epstein-Barr virus antibodies in Amer­ ican and African Burkitt’s lymphoma. Lancet ÎZ.114-116, 1973. 11. K l e i n , G . : Herpesviruses and oncogenesis. Proc. Nat. Acad. Sci. 69:1056-1064, 1972. 12. K l e i n , G., C l i f f o r d , P., K l e i n , E ., S m i t h , R. T., M i n o w a d a , J., K o u r i l s k y , F . M ., a n d B u r c h e n a l , J. H .: Membrane immunofluores­ cence reactions of Burkitt lymphoma cells from biopsy specimens and tissue cultures. J. Nat. Cancer Inst. 39:1027-1044, 1967. 13. L e v i n e , P., O ’C o n n o r , G. T., a n d B e r n a r d , C . V.: Antibodies to EB V in American Bur­ kitt’s lymphoma patients. Cancer 30:610-615, 1972.

114

H E N L E AN D H E N L E

14. N o n o y a m a , M. a n d P a g a n o , J . S.: Detection of Epstein-Barr viral genome in non-produc­ tive cells. Nature N. Biol. 233:103-106, 1971. 15. P o p e , J. H ., H o r n e , M. K., a n d W e t t e r s , E . J.: Significance of a complement-fixing antigen associated with herpes-like virus and detected in the Raji cell line. Nature 222:186187, 1969. 16. R e e d m a n , B. M. a n d K l e i n , G .: Cellular localization of an Epstein-Barr virus (E B V )associated complement-fixing antigen in non­ producer and producer lymphoblastoid cell lines. Int. J. Cancer 11:499-520, 1973. 17. R occm , G., H e w e t s o n , J., a n d H e n l e , W .: Specific neutralizing antibodies in EpsteinBarr virus-associated diseases. Int. J. Cancer 11:637-647, 1973. 18. S h o p e , T., D e c h a i r o , D ., a n d M i l l e r , G.:

Malignant lymphoma in cottontop marmosets after inoculation with Epstein-Barr virus. Proc. Nat. Acad. Sci. 70:2487-2491, 1973. 19. W o l f , H ., z u r H a u s e n , H ., a n d B e c k e r , V.: E B viral genomes in epithelial nasopharyngeal carcinoma cells. Nature N. Biol. 244:245-247, 1973. 20. z u r H a u s e n , H ., S c h u l t e -H o l t h a u s e n , H ., K l e i n , G., H e n l e , W., H e n l e , G., C l i f f o r d , P., a n d S a n t e s o n , L .: E B virus DNA in bi­ opsies of Burkitt tumors and anaplastic carci­ nomas of the nasopharynx. Nature 228:10561058, 1970. 21. z u r H a u s e n , H ., D i e h l , V., W o l f , H ., a n d S c h u l t e - H o l t h a u s e n , H . : Occurrence of Epstein-Barr virus genomes in human lympho­ blastoid cell lines. Nature N. Biol. 237:189190, 1972.