114
Journal of the Royal Society of Medicine Volume 85 February 1992
serotype 14 pneumococcal isolate,is worrying as this is the most common serotype responsible for invasive infections in children and the fourth most common in adults6. Serotype 14 strains are poorly immunogenic and are carried for the longest time after acquisition, thus allowing the greatest chance for selection of resistant strains6. The poor response to penicillin therapy of patients with meningitis caused by moderately penicillin resistant pneumococci (MIC 0.1-1 mg/I) is well recognized and results from inadequate antibiotic levels within the CSF7. In contrast, the concentrations of cefotaxime achievable within CSF across inflamed meninges are commonly 10- to 100-fold greater than the concentration required for maximum bacterial killing8. The reported clinical experience with cefotaxime in treating meningitis caused by multiply resistant pneumococci is limited but when used it has proven successful9"l, and this case supports its effectiveness. Patients with Hb SC disease are at increased risk of serious infections from encapsulated bacteria and pneumococcal vaccine is advised. The increasing incidence of antibiotic resistant pneumococci will lend support to the proposal for expanded use of pneumococcal vaccine".
2 Nair P. Incidence of decreased penicillin sensitivity of Streptococcus pneumoniae from clinical isolates. J Clin Pathol 1988; 41:720-1 3 George RC, Cooper PG, Erdman YJ. Not the first multiresistant pneumococcus in Britain. BMJ 1987;296:1206 4 Appelbaum PC, Koornhof HJ, Jacobs M, et aL Multiple antibiotic resistance of pneumococci. Morbid Mortal Weekly Rep 1977; 26:285-6 5 Gillespie SH. Aspects of pneumococcal infection including bacterial virulence, host response and vaccination. J Med Microbiol 1989;28:237-48 6 Klugman KP. Pneumococcal resistance to antibiotics. Clin Microbiol Rev 1990;3:171-96 7 Ward J. Antibiotic-resistant Streptococcus pneumoniae: clinical and epidemiological aspects. Rev Infect Dis 1981;3:254-66 8 Peretti P, Sueri L, Tosi M, et al Cefotaxime in the cerebrospinal fluid and serum in patients with purulent meningitis. J Antimicrob Chemother 1984;14(suppl B):117-23 9 Collignon PJ, Bell J, Hufton IW, Mitchell D. Meningitis caused by a penicillin- and chloramphenicol-resistant Streptococcus pneumoniae. Med J Aust 1988;149:497-8 10 Viladrich PF, Gudiol F, Linares J, et aL Characteristics and antibiotic therapy of adult meningitis due to penicillin-resistant pneumococci. Am J Med 1988;84:839-46 11 Anon. When to use the new pneumococcal vaccine. Drug Ther Bull 1990;28:31-2
References 1 Appelbaum PC. World-wide development of antibiotic resistance in pneumococci. Eur J Clin Microbiol 1987;6:367-77
(Accepted 12 February 1991)
Meeting report Viruses and malignancy Keywords: carcinogenesis; neoplasm oncagene; transformation; viruses
The meeting was organized by Dr Jennifer M Best (Reader in Virology, United Medical and Dental Schools of Guy's and St Thomas's Hospitals, London) and Professor Alan Riclinson (Professor of Cancer Studies, University of Birmingham). Professor Lutz Gissmann of the Institute for Cancer Research, Heidelberg, Germany spoke on human papillomaviruses and cervical cancer. Evidence has recently accumulated for a causative relationship between cervical cancer and certain types ofhuman papillomaviruses (HPVs), such as HPV-16, 18, 31, 331: (1) The precursor lesions of cervical cancer are induced by HPV particles as demonstrated by in vitro infection of normal tissue followed by transplantation into nude mice. (2) HPV genomes are present in a high proportion of primary and metastatic tumour biopsies and in cell lines derived therefrom. The early genes E6 and E7 are constantly transcribed and translated in the cancer cells. (3) The HPV circular genomes can be found episomally in benign (virus particle producing) genital lesions, but in the majority of cancers are integrated into the host chromosomal DNA. Integration can occur at different chromosomal locations but interruption of the HPV genome appears to be specific, usually in the 5' part of
the E2 open reading frame (ORF). This leads to isolation ofthe downstream ORFs (E4, E5, Ll, L2) from the single viral promoter region and deletions in these ORFs may be observed. In contrast, the E6 and E7 ORFs remain intact and specific mRNA is found whose synthesis is possibly facilitated by the destruction ofthe E2 ORF which codes for a transcriptional repressor. (4) The E6 and E7 ORFs were demonstrated to have biological activity in human cells in vitro. (i) Primary fibroblasts and keratinocytes can be transformed, the latter being the natural target for HPV infection. After transfection of HPV DNA, undifferentiated cells continue to grow; after a short time they become aneuploid and following numerous in vitro passages acquire the fully transformed phenotype, ie will induce tumors in nude mice. (ii) The normal pattern of epithelial differentiation, as it can be simulated in an in vitro raft system, is disturbed when HPV DNA is transfected into such epithelial cells. The histologic pattern obtained is very reminiscent of the pattern seen in HPV positive intraepithelial lesions in vivo. (iii) Expression of the E7 ORF in resting human cells leads to the induction of cellular DNA synthesis. (5) The E6/E7 ORFs seem to be involved not only in the induction of transformation (as discussed above), but also in the maintenance of the transformed state. In HPV positive cervical cell lines there is a strict correlation between the proliferation rate of the cells and the amount of E7 protein present therein. (6) A 10-fold greater prevalence and elevated titres of anti HPV-16 or HPV-18 E7 antibodies were found in sera from cervical cancer patients when
Report of Meeting of Section of Clinical Immunology & Allergy, 11 March 1991
0141-0768/92/ 020114-06/$02.00/0 © 1992 The Royal Society of Medicine
Journal of the Royal Society of Medicine Volume 85 February 1992
compared with controls, by Western blot using bacterial fusion proteins and by peptide ELISA2. It remains to be determined whether such antibodies are a useful clinical marker for early diagnosis or for follow up of the disease. Recently the mechanism of transformation by the E6/E7 proteins was clarified as it was demonstrated that they can bind and thus possibly inactivate cellular proteins (p53 and Rb) both involved in regulation of cell cycle in a negative fashion3. Because of the long period between HPV infection and the development of cervical cancer and because the development oftumors is rare, in comparison to the prevalence of HPV infection within a population, additional intrinsic mechanisms (immune regulation, cellular factors controlling HPV expression) and exogenous factors (carcinogenic substances delivered through cigarette smoking or by microorganisms) must play a role in cancer development. The mode of action of such cofactors, however, is at present only poorly understood. There is an increased prevalence of papillomavirus infections and skin cancer during immunosuppression and this was reviewed by Dr Claire Benton (University Department of Dermatology, Edinburgh, Scotland) in her presentation 'Human Papillomavirus and Skin Cancer in the Immunosuppressed Patient'. Patients immunosuppressed by disease or by drug therapy often develop cutaneous or mucocutaneous papillomavirus infections, which can be extensive and unresponsive to standard treatments. The success of the organ transplantation programme has resulted in increasing numbers of immunosuppressed allograft recipients, especially renal allograft recipients (RAR), surviving for many years post-transplantation. The consequences of prolonged immunosuppression have included the development of cancers, especially cutaneous squamous cell carcinoma4. Several factors have been incriminated in the carcinogenic process: prolonged uraemia prior to transplantation, diminished immunosurveillance, ultraviolet exposure, and the oncogenic effects of immunosuppressive drugs or viruses. Human papillomaviruses (HPV) which commonly infect allograft recipients, have been considered as possible aetiological agents in the development of cutaneous squamous cell cancers. A link between HPV and squamous cancers was first recognized in the rare autosomal recessive disease epidermodysplasia veruciformis (EV). Affected individuals show an abnormal immune response to HPV, and are infected with over 20 HPV types peculiar to this condition. Clinical signs include scaly macules and reddish-brown or pink plaques with characteristic histological changes. One third of such patients develop on light exposed sites, multiple squamous cell cancers (SCC), 90% of which contain HPV-5 or -8 DNA5. The presence of HPV-5 or 8 DNA in an individual with EV confers an increased risk of developing squamous cancers. In the EV cancers, HPV DNA is usually episomal, but its mode of action is uncertain. However, both HPV-5 and HPV-8 DNA are capable of transforming rodent cell lines in vitro. Clinical, histological and virological evidence now suggests a possible role for HPV in cutaneous carcinogenesis in the immunosuppressed. Long-term allograft recipients are at increased risk of developing both cutaneous warts and squamous cancers, often
multiple, especially if ultraviolet exposure has been excessive. There appears to be an overlapping spectrum of disease from viral wart and verrucous keratosis to dysplastic keratosis and in-situ or invasive squamous cancer. This clinical spectrum is reflected histologically with epidermal dysplasia seen in viral warts, and viral features found even in invasive squamous cancers. Plaques, clinically and histologically typical of epidermodysplasia verruciformis, and containing HPV-5 or -8 DNA have also been observed in several female RAR6. In a study in Edinburgh, EV types of HPV DNA were detected in 13 of a group of 72 RAR; and HPV-5/8 DNA was found in 60% of squamous cancers examined as well as 18% of warts and keratoses7. The presence of these HPV types in an immunosuppressed individual was significantly more frequent in those patients with cutaneous cancers, and appeared to predispose to their future development as is the case in epidermodysplasia verruciformis. Detection of EV types of HPV DNA in cutaneous cancers is, on its own, insufficient to prove a definite casual effect: further investigation is required to elucidate the molecular mechanis involved. The use of the polymerase chain reaction will also assist in assessing the prevalence of EV types of HPV in both inmunocompetent and immunosuppssed populations as serological evidence suggests that HPV-8 may be more widespread than was previously thought8. Dr Benton recommended that renal allograft recipients should be monitored regularly for cutaneous infections and tumours, and should be advised to avoid excessive sun exposure. Professor Beverly Griffin (Royal Postgraduate Medical School, London) spoke on Epstein-Barr virus and nasopharyngeal carcinoma. In Britain, where lung (for men) and breast (for women) cancers dominate the statistics, nasopharyngeal carcinoma (NPC) is rarely observed and generally does not appear among the cancer incidence figures9. In Southern China and other parts of Asia, however, the situation is quite different and the incidence of NPC is as high as 15-30 cases per 100 000 people per year, placing it on the high incidence list in those areas of the world'0. Its restricted location even in countries where NPCs are commonly seen (in China, it is often called Guandong (Canton) tumour) is still, however, unexplained. It is essentially a tumour of an ageing population in Asia but, interestingly, is also observed in relatively high numbers in North Africa, where the distribution has been observed to be bimodal, with 15% of the cases developing between 10 and 25 years of agel1. The association of the human herpesvirus, Epstein-Barr virus (or EBV), with NPC was made in the 1960s almost by chance. Serum samples sent by Mr Peter Clifford, an ENT Surgeon in Nairobi, to New York as controls for a study on Burkitt's lymphoma (BL, another tumour associated with EBV, discussed by G Klein elsewhere in this article) also included among them sera from patients with NPC. Old and collaborators'2 recognized antibody identities in the samples and were able to relate them to the presence of EBV in both cases. Table 1 shows that the EB viral association is almost the only property shared by these two quite distinct diseases. During nearly 20 years that followed Old's discovery, many excellent scientific groups drawn from the fields
115
116
Journal of the Royal Society of Medicine Volume 85 February 1992
Table 1. The major malignancies associated with EBV: Burkitt's lymphoma (BL) and nasopharyngeal carcinoma (NPC)
NPC
Endemic BL
Face, abdomen, CNS lymph node, bone marrow Cell type Immature B-cell Africa (West Nile) Geographical bias Papua New Guinea Sex and age Male>female, 6-14 years Annual incidence 50-100/million/yr (Uganda) Treatment Chemotherapy Chromosomal abnormalities t (8; 14), t (8; 2), t (8; 22), Other Implicated cellular oncogenes c-myc EA-R (IgG) dominant Elevated virus specific antibodies to early antigens Cell culture Easy Suggested co-factor Malaria Tumour site
'head and neck' Poorly differentiated epithelium Southern China, Hong Kong, Hawaii, Alaska, Singapore, North and East Africa Male>female, 40-65 years 470/million/yr (Hong Kong) Radiotherapy Abnormal 3* None identified EA-D (IgA) dominant
Difficult Carcinogens (salted fish)
*From two tumours
of virology, immunology and oncology, focused their attention on BL and the role played by EBV in the genesis of that tumour, more or less to the exclusion of NPC. Studies on the latter largely attracted epidemiologists and clinicians interested in discovering the genetic or lifestyle differences that might account for the restricted location and high incidence of NPC in certain parts of the world. It was obvious that factors other than infection with EBV, which is ubiquitous among the world's population, must be contributing to disease, and nitrosamines in salted fish, a staple diet in high incidence areas, received much attention13 as a 'cofactor' in disease, but the relationship remains, to date, unproved. The last few years have, however, seen an awakening interest in NPC, particularly among virologists and molecular biologists, and a greater exchange of information and materials between Eastern and Western parts of the world is greatly contributing to increased knowledge. The data presented in Table 1 also suggest that if EBV is a principal contributor to both BL and NPC, there must either exist different pathological strains of the virus (as in the case of herpes simplex, where two distinct types, HSV-1 and HSV-2, exist) or else the host cells, themselves, must be differentially permissive to viral gene expression, exerting their own control in the symbiotic relationship with the virus. With regard to the former, whereas there are variations among viral isolates, there are no convincing data relating to pathologically variant strains. Until recently there have been no data relevant to the latter. Our own interest in NPC began in the early 1980s when, with molecular techniques, we dissected the viral DNA into large overlapping fragments and used these to ascertain whether functions could be identified that altered the growth properties in culture of epithelial cells. One, among eight EBV DNA fragments tested in this way, had unusual properties. That is, in a mixture of primary primate cells (human or monkey) one specific fragment among them consistently stimulated epithelial cells in the population to proliferate indefinitely in culture and some of these became 'immortalized', whereas fibroblasts in the same population were unaffected in their growth'4 5. It was notable that this fragment in question contained none of the viral genes [the family of EBV nuclear antigens (EBNAs) or a membrane protein, designated latent membrane
protein (LMP)] that had hitherto been associated with viral stimulation and immortalization of Blymphocytes. Pagano and collaborators'6 analysed EBV transcriptional expression in several biopsies from American NPC patients and found messenger RNA from limited, but not wholly consistent, regions of the viral genome. Thus, the data at this stage pointed to cell specific EBV genes being expressed. This notion was considerably strengthened when we, with T Tursz and co-workers in Paris, began to analyse viral gene expression in detail in an NPC tumour from North Africa that had been established to grow in nude mice. Our data17 showed that the region we had already been studying with regard to cell growth stimulation was highly expressed, whereas other regions of the viral genome were essentially silent. Confirmatory evidence for these findings are now coming out of other laboratories, including one in Hong Kong focusing on EB viral gene expression in Asian NPCs (H-L Chen, M Lung and M H Ng, personal communication). The data being accumulated are consistent, not only with the notion that where EBV is concerned, the North African and Asian tumours appear to be similar, but virally important genes in the epithelial cell environment may not be totally identical to those related to the B-cell disease, BL. With the increasing attention now being given to NPC, we should soon know considerably more about the viral involvement in this disease. (Ironically, this change in the focus of interest has resulted, in some circles at least, in questions being raised regarding any role for EBV in BL. This concern is strengthened by the fact that there seems to be no involvement of EBV in non-endemic BL.) It has been known for some time that prior to the appearance of disease, IgA titres to EBV genes, such as the viral capsid antigen, VCA, rise sharply. In a wide-ranging serological screening of Chinese in NPC high incidence areas, people at risk of developing the disease were identified; follow-up studies of these showed that high IgA/VCA titres preceded clinical symptoms by 6-41 months'8. More recently, 130 asymptomatic patients from 6054 individuals with high EBV antibody titres were randomly recruited and examined by fibreoptic endoscopy'9. From among these, seven cases of NPC were detected. The tumour was confirmed as largely confined to the pharyngeal recess. These studies support the involvement of EBV
Journal of the Royal Society of Medicine Volume 85 February 1992
in the disease and at the same time point to the importance of the development of simple screening procedures for identifying people either carrying, or at risk of developing, this cancer considerably before it becomes evident by conventional means. Early patient management, as in most other human cancers, affords a much higher treatment success rate. Prognostic methods are available for NPC but, as we learn more about EBV gene expression in epithelial cells, can undoubtedly be improved. Dr George Klein (Department of Tumor Biology, Karolinska Institutet, Stockholm, Sweden) spoke on 'The Relative Role of the EBV Transformation and the Ig/myc Translocation in the Genesis of Burkitt's Lymphoma'. Translocation of the c-myc protooncogene to an immunoglobulin locus is a rate-limiting step in the genesis of three B-cell derived tumours: Burkitt lymphoma (BL) in humans, mouse plasmacytoma (MPG) and rat immunocytoma (RIO)20. The translocation must represent a rate-limiting step in the genesis of all three tumours. Its consequences have best been analysed in BL. They involve a nonimmunological and an immunological component. The former acts by preventing the B-cell from leaving the cycling compartment and entering the resting stage when programmed to do so. The latter acts by the down-regulation of certain HLA class I polymorphic specificities, adhesion molecules and EBV-encoded proteins. According to Dr Klein's interpretation, the translocation fixes the BL cell in a phenotypic window that can be referred to as 'a resting cell that is not resting'. The linking of c-myc to immunoglobulin sequences in a B-cell leads to constitutive myc expression. In spite of its switch to a 'resting phonotype', the cell is unable to leave the cycling compartment2l. Normally, EBV-carrying B-blasts would face immune controls. They can only proliferate in immunodefectives, as in XLP or transplant lymphoma. Due to its 'resting' rather than B-blast phenotype and the correlated defective expression of certain EBV, HLA and adhesion molecules, however, the BL cell is not controlled by the immune response. The down regulation of EBNA 2-6 and latent membrane protein (LMP) in the BL cell were also considered in relation to viral latency. The exclusive expression of EBNA 1 in the BL cell may reflect the adaptation of the virus to prolonged persistence in long lived resting B-cells.
'Hepatitis B viruses and hepatocellular carcinoma' was the topic presented by Dr Marie Annick Buendia (Unite de Recombinaison et Expression Genetique (Inserm U163), Institut Pasteur, Paris, France). Primary hepatocellular carcinoma (HCC), one ofthe most common cancers in many parts of the world, also shows seroepidemiological association with a viral infection: hepatitis B virus (HBV) has been clearly identified as a major cause of HOC. During the last decade, extensive studies, mainly focused on the HBV life cycle, the function of HBV proteins, the regulation of viral gene expression and the mechanisms of viral DNA integration into host cell DNA, have sustained different hypotheses on the viral contribution to carcinogenesis. Arguments favouring both direct and indirect mechanisms have accumulated recently. Integrated HBV sequences,
found in most HCCs examined, might play a part in carcinogenesis, either directly, through insertional activation of cellular genes, or indirectly, by altering the cell growth control in trans. In two independent tumours, HBV insertion sites have been identified as the receptor B for retinoic acid and the cyclin A genes22'23. In other tumours, truncated products of integrated viral X gene and PreS2/S region have been shown to harbour transcriptional trans-activation properties24'25. More indirect mechanisms, involving prolonged liver cell injury mediated by the antiviral immune response, and continuous hepatocyte regeneration, have also been proposed. Accordingly, overexpression of unsecretable, large HBV envelope protein in transgenic mice has been correlated with liver necrosis, inflammation and eventually HCC26. Much less attention, however, has been paid to the underlying cellular mechanisms leading to malignant transformation in response to viral induction. To date, no unifying model, accounting for the contribution of viral and cellular factors to liver oncogenesis, has been proposed. In contrast, studies of woodchuck liver tumours associated with chronic infection with woodchuck hepatitis virus (WHV) have revealed a more simple pattern. WHV, a virus closely related to HBV, is used as a model for liver disease and hepatocarcinogenesis in humans. More than 80% of carrier animals succumb to HCC after 4 years; moreover, the oncogenic properties of WHV, in the absence of carcinogenic cofactors, have been demonstrated by the induction of HCC in experimentally infected, newborn woodchucks. In a recent analysis of 30 woodchuck liver tumours, we have observed frequent activation of myc family genes, either by insertion of WHV DNA or genetic rearrangement, or by a trans-acting mechanism. In 30% of the tumours analysed, viral integration occurred within either c-myc, N-myc or a newly identified N-myc retroposon27 25. We have shown that comparable mechanisms, leading to deregulated expression of myc genes, operate in the development of tumours induced by hepatitis virus or by nonacute retroviruses in rodents. The contribution of the mutated myc genes to hepatocyte transformation has been demonstrated by the induction of primary liver tumours in transgenic mice carrying a woodchuck myc gene activated by hepadnavirus insertion. These results show that insertional activation of cellular proto-oncogenes by WHV DNA is a common event in woodchuck HCC, and implicate myc family genes in the genesis of most of these tumours.
Professor Howard Thomas (Department of Medicine, St Mary's Hospital, Medical School) also spoke about the association of HBV with HCC. Dr James Neil (Beatson Institute for 'Cancer Research, Glasgow) introduced the session on retroviruses, with his talk on 'Molecular pathogenesis of feline leukaemia vireus. Genetic damage arising from proviral insertion and recombination is now familiar as an explanation of the oncogeneity of type-C retroviruses. Retroviral 'tagging' of host cell genes by these processes underlies their pre-eminent role in elucidating the genetic basis of cancer. Feline leukaemia virus (FeLV) has been one ofthree highly productive models in this field. However, in natural FeLV infection, tumours
117
118
Journal of the Royal Society of Medicine Volume 85 February 1992
develop only rarely, and after a long latent period. To understand the basis of oncogenesis by these viral agents clearly requires a deeper knowledge of the virus-host interaction. Although the infectious, horizontally spread form of FeLV (FeLV-A) is remarkably conserved, its replication in vivo generates diverse variants which are associated with specific acute diseases. Thus, acutely pathogenic isolates of FeLV are invariably mixed populations of the replication-competent FeLVA and one or more variant genome. The reproduction of acute disease requires both the common and variant genomes to be present. Although FeLV variants are important mediators of acute disease they may thereby hamper virus dissemination and are not themselves efficiently transmitted under natural
conditions29. The variants include envelope gene mutants (FeLVC, FeLV-FAIDS) and recombinants with endogenous proviruses (FeLV-B) some of which have expanded host range and increased pathogenic potential30. Another site at which FeLV evolves from the prototype is in the proviral long terminal repeats where duplications of transcriptional enhancer elements are often present in tumour derived proviruses31. A third class of FeLV variant is represented by oncogene containing recombinants. However, acutely oncogenic variants appear to arise only rarely since neoplastic disease is an uncommon outcome of natural virus infection. Our recent work has focused mainly on the aetiology of the most frequently observed FeLV-associated neoplasm, thymic lymphosarcoma. Many of these tumours involve activation of the c-myc gene, either by proviral insertion or transduction. However, even where potent FeLV v-myc oncogenes are present, the tumours are of clonal origin and may have an immortal, factor-independent growth phenotype which has not so far been reproduced by in vitro infection32. It seems, therefore that other oncogenic events mediate progression of tumours initiated by an activated c-myc allele. By analysing proviral integration sites in tumours induced by FeLV-myc recombinant viruses, we have recently identified two common regions of
proviral insertion, designated fit-1 and fit-2. Gene fragments from these loci, which have no obvious homologue in gene sequence data-banks, detect two proviral integration clusters in four out of 13 independent tumours (H Tsujimoto and J C Neil, unpublished results). In one tumour (T17) a second transductant was present in addition to the myccontaining FeLV in the form of multiple proviruses carrying the complete coding sequence of a T-cell antigen receptor ,3-chain gene33. Although experimental infection with this recombinant has failed to demonstrate oncogenic potential for secondary hosts, an antigen or MHC-restricted oncogenic function remains plausible although difficult to evaluate in the outbred cat. Our current working model for the induction of thymic lymphosarcoma by FeLV is that accidents of virus replication disrupt the sensitive process of thymic education, whereby potentially autoreactive T-lymphocytes are normally recognized and induced to undergo programmed cell death. To test this as a general hypothesis in a more defined system, we have produced transgenic mouse lines in which the oncogenicity of c-myc is directed to the T-cell lineage by linking it to a dominant control region from the
human CD2 gene. These mice develop sporadic thymic tumours which can be greatly accelerated by neonatal infection with Moloney murine leukaemia virus (M Stewart, E Cameron, D Onions and J C Neil, unpublished). We believe that this will be an invaluable system in which to test the roles of the T-cell receptor repertoire and of loci identified by proviral insertion. Professor Jonathan Weber (Department of Genitourinary Medicine and Communicable Diseases, St Mary's Hospital) spoke on 'Human T-cell lymphotropic viruses'. Dr Valerie Beral (ICRF Cancer Epidemiology Unit, Oxford) spoke on 'HIV, AIDS and cancer'. In Europe and the USA one in five people with AIDS also have cancer, Kaposi's sarcoma and non-Hodgkin's lymphoma are the most frequent malignancies which occur in HIV-infected individuals. These cancers are rare in H1V uninfected people, except in other immunosuppressed groups such as transplant recipients. Thus the HIV-associated immunosuppression plays a major role in the development of these tumours. It is widely believed that the cancers which occur in the immunosuppressed have a viral aetiology. Kaposi's sarcoma is nevertheless 300 times as common in AIDS patients than in other immunosuppressed groups. It has been reported in 20% of all patients with AIDS in Britain, and is almost entirely confined to men who acquired the HIV by homosexual contact and, to a lesser extent, to heterosexuals who acquired their HIV infection in Africa3'435. Kaposi's sarcoma is, however, rare in patients with AIDS who acquired the HIV by non-sexual routes. The evidence to hand suggests that Kaposi's sarcoma in people with AIDS is caused by an as yet unidentified infectious agent and that the immunosuppression associated with HIV infection leads to clinical Kaposi's sarcoma in an otherwise asymptomatic carrier of the agent. That the risk of Kaposi's sarcoma is much higher in those who acquired their HIV sexually than by other means suggests that a major mode of transmission of the agent is sexual. Furthermore the higher risk of Kaposi's sarcoma in homosexuals than heterosexuals suggests that some aspect of homosexual behaviour facilitates the transmission of the agent. Non-Hodgkin's lymphoma occurs in about 3% of subjects with AIDS, irrespective of their means of acquiring HIV infection36. The lymphomas are clinically atypical, often arising in extrandodal sites. Primary lymphoma of the brain, which is normally exceedingly rare, accounts for about one fifth of the lymphomas which develop in patients with AIDS. The lymphomas tend to be high-grade lesions, most being classified as immunoblastic lymphomas. Burkitt's lymphoma also occurs in patients with AIDS and this is surprising, since Burkitt's lymphomas do not arise in other immunosuppressed individuals, whereas immunoblastic lymphomas do. Apart from the role of immunosuppression, the reason why lymphomas develop in people with AIDS is unclear. Infectious agents are believed to be involved and EBV may well play a part. EBV is commonly associated with African Burkitt's lymphoma not in other forms of lymphomas which arise in the immunosuppressed. Its association with HIV-associated lymphomas is not so clear. Whether the risk of other cancers is increased in HIV-infected individuals is still unknown.
Journal of the Royal Society of Medicine Volume 85 February 1992
Likely candidates are cancers with a known or suspected viral origin, such as hepatoma, cervical cancer and Hodgkin's disease, since the HIV-induced immunosuppression may well accelerate the process of viral carcinogenesis.
Jennifer M Best Department of Virology St Thomas' Hospital Lambeth Palace Road London SEl 7EH References 1 zur Hausen H. Papillomaviruses in anogenital cancer as a model to understanding the role of viruses in human cancers. Cancer Res 1989;49:4677-81 2 Jochmus-Kudielka I, Schneider A, Braun R, et al. Antibodies against the human papillomavirus type 16 early proteins in human sera: correlation of anti-E7 reactivity with cervical cancer. J Natl Cancer Inst 1989;81:1698-1704 3 Werness BA, Levine AJ, Howley PM. Association of human papillomavirus types 16 and 18 E6 proteins with p53. Science 1990;248:76-9 4 Kinlen LJ, Sheil AGR, Peto J, Doll R. Collaborative United Kingdom Australasian study of cancer in patients treated with immunosuppressive drugs. BMJ 1979;ii:1461-6 5 Orth G. Epidermodysplasia verruciformis. In: Salzman NP, Howley PM, eds. The Papovaviridae. The papillomaviruses. New York: Plenum, 199-243 6 Rudlinger R, Smith IW, Bunney MH, Hunter JAA. Human papilloma virus infections in a group of renal transplant recipients. Br J Dermatol 1986;115:681-92 7 Barr BBB, Benton EC, McLaren K, et al. Human papilloma virus infection and skin cancer in renal allograft recipients. Lancet 1989;i:124-9 8 Steger G, Olszewsku M, Stockfleth E, Pfister H. Prevalence of antibodies to human papillomavirus type 8 in human sera. J Virol 1990;64:4399-406 9 Cancer statistics: registration, England and Wales 1984. HMSO, London 1988; Scottish Health Statistics 1986/7; Statistics and Research Bulletin, SR1/87 DHSS Belfast 1988. (Taken from report, Cancer Research Campaign, 1989) 10 Zeng Y, Ou B. Etiology and pathogenesis of nasopharyngeal carcinoma. China: The People's Publishing House, 1985 11 de-The G. Role of Epstein-Barr virus in human diseases: Infectious mononucleosis, Burkitt's lymphoma, and nasopharyngeal carcinoma. In: Klein G, ed. Viral oncology. New York: Raven Press, 1980:769-97 12 Old LJ, Boyse EA, Oettegen HF, et al. Precipitation antibody in human serum to an antigen present in cultured Burkitt's lymphoma cells. Proc Natl Acad Sci USA 1966;56:1699-704 13 Huang DP, Ho JHC, Saw D, Teoh TB. Carcinoma ofthe nasal and paranasal regions in rats fed Cantonese salted marine fish. In: de-The G, Ito Y, eds. Nasopharyngeal carcinoma; etiology and Control. Lyon: IARC, 1978:315-18 14 Griffin B, Karran L. Immortalization of monkey epithelial cells by specific fragments of Epstein-Barr virus DNA. Nature 1984;309:78-82 15 Karran L, Teo CG, King D, et aL Establishment of immortalized primate epithelial cells with sub-genomic EBV DNA. Int J Cancer 1990;46:763-72 16 Raab-Traub N, Hood R, Yang C-S, Henry B, Pagano JS. Epstein-Barr virus transcription in nasopharyngeal carcinoma. J Virol 1983;48:580-90
17 Hitt MM, Allday MJ, Hara T, Karran L, Jones MD, Bussons P, Tursz T, Ernberg I, Griffin BE. EBV gene expression in an NPC-related tumour. EMBO J 1989; 8:2639-51 18 Zeng Y, Shen SJ, Deng H, et al. Early nasopharyngeal carcinoma among IgA/VCA antibody positive individuals detected by anti-complement immunoenzymatic medicine. Chinese Med J 1984;97:155-7 19 Sham JST, Wei W, Zong Y-S, et al. Detection of subclinical nasopharyngeal carcinoma by fibreoptic endoscopy and multiple biopsy. Lancet 1990;335:371-3 20 Klein G. Viral latency and transformation: The strategy of Epstein-Barr virus. Cell 1989;58:5-8 21 Klein G. Multiple phenotypic consequences of the Ig/myc translocation in B-cell-derived tumors. Genes, Chromosomes Cancer 1989;1:3-8 22 Dejean A, Bougueleret L, Crzeschik KH, Tiollais P. Hepatitis B virus DNA integration in a sequence homologous to v-erb-A and steroid receptor genes in a hepatocellular carcinoma. Nature 1986;322:70-2 23 Wang J, Chenivesse X, Henglein B, Brechot C. Hepatitis B virus integration in a cyclin A gene in a hepatocellular carcinoma. Nature 1990;343:555-7 24 Wollersheim M, Debelka U, Hofschneider PH. A transactivating function encoded in the hepatitis B virus X gene is conserved in the integrated state. Oncogene 1988;3:545-54 25 Kekule AS, Lauer U, Meyer M, Caselman WH, Hofschneider PH, Koshy R. The pre S2/S region of integrated hepatitis B virus DNA encodes a transcriptional transactivator. Nature 1990;342:457-61 26 Chisari FV, Klopchin K, Moriyama T, et aL Molecular pathogenesis of hepatocellular carcinoma in hepatitis B virus transgenic mice. Cell 1989;59:1145-56 27 Hsu TY, Moroy T, Etiemble J, et aL Activation of c-myc by Woodchuck hepatitis virus insertion in hepatocellular carcinoma. Cell 1988;55:627-35 28 Fourel G, Trepo C, Bougueleret L, et al. Frequent activation of N-myc genes of hepadnavirus insertion in woodchuck liver tumours. Nature 1990;347:294-8 29 Neil JC, Fulton R, Rigby M, Stewart M. Evolution of pathogenic and oncogenic variants of feline leukaemia virus. Curr Top Microbiol Immunol 1991, in press 30 Stewart MA, Warnock M, Wheeler A, et al. Nucleotide sequences of a feline leukemia virus subgroup A envelope gene and long terminal repeat and evidence for the recombinational origin of subgroup B viruses. J Virol 1986;58:825-34 31 Fulton R, Plumb M, Shield L, Neil JC. Structural diversity and nuclear protein binding sites in the long terminal repeats of feline leukaemia virus. J Virol 1990;64:1675-82 32 Onions D, Lees G, Forrest D, Neil JC. Recombinant feline viruses containing the myc gene rapidly produce clonal tumours expressing T-cell antigen receptor gene transcripts. Int J Cancer 1987;40:40-5 33 Fulton R, Forrest D, McFarlane R, Onions D, Neil JC. Retroviral transduction of T-cell antigen receptor Bchain and myc genes. Nature 1987;326:190-4 34 Beral V, Peterman TA, Berkelman R, Jaffe HW. Kaposi's sarcoma among persons with AIDS: a sexually transmitted infection? Lancet 1990;335:123-8 35 Beral V, Bull D, Jaffe H, Evans B, Gill N, Tillett H, Swerdlow AJ. Is risk of Kaposi's sarcoma in AIDS patients in Britain increased if sexual partners come from United States or Africa? BMJ 1991;302:624-5 36 Beral V, Peterman T, Berkelman R, Jaffe H. AIDSassociated non-Hodgkin's lymphoma. Lancet 1991; 337:805-9
119
Journal of the Royal Society of Medicine Volume 85 February 1992
serotype 14 pneumococcal isolate,is worrying as this is the most common serotype responsible for invasive infections in children and the fourth most common in adults6. Serotype 14 strains are poorly immunogenic and are carried for the longest time after acquisition, thus allowing the greatest chance for selection of resistant strains6. The poor response to penicillin therapy of patients with meningitis caused by moderately penicillin resistant pneumococci (MIC 0.1-1 mg/I) is well recognized and results from inadequate antibiotic levels within the CSF7. In contrast, the concentrations of cefotaxime achievable within CSF across inflamed meninges are commonly 10- to 100-fold greater than the concentration required for maximum bacterial killing8. The reported clinical experience with cefotaxime in treating meningitis caused by multiply resistant pneumococci is limited but when used it has proven successful9"l, and this case supports its effectiveness. Patients with Hb SC disease are at increased risk of serious infections from encapsulated bacteria and pneumococcal vaccine is advised. The increasing incidence of antibiotic resistant pneumococci will lend support to the proposal for expanded use of pneumococcal vaccine".
2 Nair P. Incidence of decreased penicillin sensitivity of Streptococcus pneumoniae from clinical isolates. J Clin Pathol 1988; 41:720-1 3 George RC, Cooper PG, Erdman YJ. Not the first multiresistant pneumococcus in Britain. BMJ 1987;296:1206 4 Appelbaum PC, Koornhof HJ, Jacobs M, et aL Multiple antibiotic resistance of pneumococci. Morbid Mortal Weekly Rep 1977; 26:285-6 5 Gillespie SH. Aspects of pneumococcal infection including bacterial virulence, host response and vaccination. J Med Microbiol 1989;28:237-48 6 Klugman KP. Pneumococcal resistance to antibiotics. Clin Microbiol Rev 1990;3:171-96 7 Ward J. Antibiotic-resistant Streptococcus pneumoniae: clinical and epidemiological aspects. Rev Infect Dis 1981;3:254-66 8 Peretti P, Sueri L, Tosi M, et al Cefotaxime in the cerebrospinal fluid and serum in patients with purulent meningitis. J Antimicrob Chemother 1984;14(suppl B):117-23 9 Collignon PJ, Bell J, Hufton IW, Mitchell D. Meningitis caused by a penicillin- and chloramphenicol-resistant Streptococcus pneumoniae. Med J Aust 1988;149:497-8 10 Viladrich PF, Gudiol F, Linares J, et aL Characteristics and antibiotic therapy of adult meningitis due to penicillin-resistant pneumococci. Am J Med 1988;84:839-46 11 Anon. When to use the new pneumococcal vaccine. Drug Ther Bull 1990;28:31-2
References 1 Appelbaum PC. World-wide development of antibiotic resistance in pneumococci. Eur J Clin Microbiol 1987;6:367-77
(Accepted 12 February 1991)
Meeting report Viruses and malignancy Keywords: carcinogenesis; neoplasm oncagene; transformation; viruses
The meeting was organized by Dr Jennifer M Best (Reader in Virology, United Medical and Dental Schools of Guy's and St Thomas's Hospitals, London) and Professor Alan Riclinson (Professor of Cancer Studies, University of Birmingham). Professor Lutz Gissmann of the Institute for Cancer Research, Heidelberg, Germany spoke on human papillomaviruses and cervical cancer. Evidence has recently accumulated for a causative relationship between cervical cancer and certain types ofhuman papillomaviruses (HPVs), such as HPV-16, 18, 31, 331: (1) The precursor lesions of cervical cancer are induced by HPV particles as demonstrated by in vitro infection of normal tissue followed by transplantation into nude mice. (2) HPV genomes are present in a high proportion of primary and metastatic tumour biopsies and in cell lines derived therefrom. The early genes E6 and E7 are constantly transcribed and translated in the cancer cells. (3) The HPV circular genomes can be found episomally in benign (virus particle producing) genital lesions, but in the majority of cancers are integrated into the host chromosomal DNA. Integration can occur at different chromosomal locations but interruption of the HPV genome appears to be specific, usually in the 5' part of
the E2 open reading frame (ORF). This leads to isolation ofthe downstream ORFs (E4, E5, Ll, L2) from the single viral promoter region and deletions in these ORFs may be observed. In contrast, the E6 and E7 ORFs remain intact and specific mRNA is found whose synthesis is possibly facilitated by the destruction ofthe E2 ORF which codes for a transcriptional repressor. (4) The E6 and E7 ORFs were demonstrated to have biological activity in human cells in vitro. (i) Primary fibroblasts and keratinocytes can be transformed, the latter being the natural target for HPV infection. After transfection of HPV DNA, undifferentiated cells continue to grow; after a short time they become aneuploid and following numerous in vitro passages acquire the fully transformed phenotype, ie will induce tumors in nude mice. (ii) The normal pattern of epithelial differentiation, as it can be simulated in an in vitro raft system, is disturbed when HPV DNA is transfected into such epithelial cells. The histologic pattern obtained is very reminiscent of the pattern seen in HPV positive intraepithelial lesions in vivo. (iii) Expression of the E7 ORF in resting human cells leads to the induction of cellular DNA synthesis. (5) The E6/E7 ORFs seem to be involved not only in the induction of transformation (as discussed above), but also in the maintenance of the transformed state. In HPV positive cervical cell lines there is a strict correlation between the proliferation rate of the cells and the amount of E7 protein present therein. (6) A 10-fold greater prevalence and elevated titres of anti HPV-16 or HPV-18 E7 antibodies were found in sera from cervical cancer patients when
Report of Meeting of Section of Clinical Immunology & Allergy, 11 March 1991
0141-0768/92/ 020114-06/$02.00/0 © 1992 The Royal Society of Medicine
Journal of the Royal Society of Medicine Volume 85 February 1992
compared with controls, by Western blot using bacterial fusion proteins and by peptide ELISA2. It remains to be determined whether such antibodies are a useful clinical marker for early diagnosis or for follow up of the disease. Recently the mechanism of transformation by the E6/E7 proteins was clarified as it was demonstrated that they can bind and thus possibly inactivate cellular proteins (p53 and Rb) both involved in regulation of cell cycle in a negative fashion3. Because of the long period between HPV infection and the development of cervical cancer and because the development oftumors is rare, in comparison to the prevalence of HPV infection within a population, additional intrinsic mechanisms (immune regulation, cellular factors controlling HPV expression) and exogenous factors (carcinogenic substances delivered through cigarette smoking or by microorganisms) must play a role in cancer development. The mode of action of such cofactors, however, is at present only poorly understood. There is an increased prevalence of papillomavirus infections and skin cancer during immunosuppression and this was reviewed by Dr Claire Benton (University Department of Dermatology, Edinburgh, Scotland) in her presentation 'Human Papillomavirus and Skin Cancer in the Immunosuppressed Patient'. Patients immunosuppressed by disease or by drug therapy often develop cutaneous or mucocutaneous papillomavirus infections, which can be extensive and unresponsive to standard treatments. The success of the organ transplantation programme has resulted in increasing numbers of immunosuppressed allograft recipients, especially renal allograft recipients (RAR), surviving for many years post-transplantation. The consequences of prolonged immunosuppression have included the development of cancers, especially cutaneous squamous cell carcinoma4. Several factors have been incriminated in the carcinogenic process: prolonged uraemia prior to transplantation, diminished immunosurveillance, ultraviolet exposure, and the oncogenic effects of immunosuppressive drugs or viruses. Human papillomaviruses (HPV) which commonly infect allograft recipients, have been considered as possible aetiological agents in the development of cutaneous squamous cell cancers. A link between HPV and squamous cancers was first recognized in the rare autosomal recessive disease epidermodysplasia veruciformis (EV). Affected individuals show an abnormal immune response to HPV, and are infected with over 20 HPV types peculiar to this condition. Clinical signs include scaly macules and reddish-brown or pink plaques with characteristic histological changes. One third of such patients develop on light exposed sites, multiple squamous cell cancers (SCC), 90% of which contain HPV-5 or -8 DNA5. The presence of HPV-5 or 8 DNA in an individual with EV confers an increased risk of developing squamous cancers. In the EV cancers, HPV DNA is usually episomal, but its mode of action is uncertain. However, both HPV-5 and HPV-8 DNA are capable of transforming rodent cell lines in vitro. Clinical, histological and virological evidence now suggests a possible role for HPV in cutaneous carcinogenesis in the immunosuppressed. Long-term allograft recipients are at increased risk of developing both cutaneous warts and squamous cancers, often
multiple, especially if ultraviolet exposure has been excessive. There appears to be an overlapping spectrum of disease from viral wart and verrucous keratosis to dysplastic keratosis and in-situ or invasive squamous cancer. This clinical spectrum is reflected histologically with epidermal dysplasia seen in viral warts, and viral features found even in invasive squamous cancers. Plaques, clinically and histologically typical of epidermodysplasia verruciformis, and containing HPV-5 or -8 DNA have also been observed in several female RAR6. In a study in Edinburgh, EV types of HPV DNA were detected in 13 of a group of 72 RAR; and HPV-5/8 DNA was found in 60% of squamous cancers examined as well as 18% of warts and keratoses7. The presence of these HPV types in an immunosuppressed individual was significantly more frequent in those patients with cutaneous cancers, and appeared to predispose to their future development as is the case in epidermodysplasia verruciformis. Detection of EV types of HPV DNA in cutaneous cancers is, on its own, insufficient to prove a definite casual effect: further investigation is required to elucidate the molecular mechanis involved. The use of the polymerase chain reaction will also assist in assessing the prevalence of EV types of HPV in both inmunocompetent and immunosuppssed populations as serological evidence suggests that HPV-8 may be more widespread than was previously thought8. Dr Benton recommended that renal allograft recipients should be monitored regularly for cutaneous infections and tumours, and should be advised to avoid excessive sun exposure. Professor Beverly Griffin (Royal Postgraduate Medical School, London) spoke on Epstein-Barr virus and nasopharyngeal carcinoma. In Britain, where lung (for men) and breast (for women) cancers dominate the statistics, nasopharyngeal carcinoma (NPC) is rarely observed and generally does not appear among the cancer incidence figures9. In Southern China and other parts of Asia, however, the situation is quite different and the incidence of NPC is as high as 15-30 cases per 100 000 people per year, placing it on the high incidence list in those areas of the world'0. Its restricted location even in countries where NPCs are commonly seen (in China, it is often called Guandong (Canton) tumour) is still, however, unexplained. It is essentially a tumour of an ageing population in Asia but, interestingly, is also observed in relatively high numbers in North Africa, where the distribution has been observed to be bimodal, with 15% of the cases developing between 10 and 25 years of agel1. The association of the human herpesvirus, Epstein-Barr virus (or EBV), with NPC was made in the 1960s almost by chance. Serum samples sent by Mr Peter Clifford, an ENT Surgeon in Nairobi, to New York as controls for a study on Burkitt's lymphoma (BL, another tumour associated with EBV, discussed by G Klein elsewhere in this article) also included among them sera from patients with NPC. Old and collaborators'2 recognized antibody identities in the samples and were able to relate them to the presence of EBV in both cases. Table 1 shows that the EB viral association is almost the only property shared by these two quite distinct diseases. During nearly 20 years that followed Old's discovery, many excellent scientific groups drawn from the fields
115
116
Journal of the Royal Society of Medicine Volume 85 February 1992
Table 1. The major malignancies associated with EBV: Burkitt's lymphoma (BL) and nasopharyngeal carcinoma (NPC)
NPC
Endemic BL
Face, abdomen, CNS lymph node, bone marrow Cell type Immature B-cell Africa (West Nile) Geographical bias Papua New Guinea Sex and age Male>female, 6-14 years Annual incidence 50-100/million/yr (Uganda) Treatment Chemotherapy Chromosomal abnormalities t (8; 14), t (8; 2), t (8; 22), Other Implicated cellular oncogenes c-myc EA-R (IgG) dominant Elevated virus specific antibodies to early antigens Cell culture Easy Suggested co-factor Malaria Tumour site
'head and neck' Poorly differentiated epithelium Southern China, Hong Kong, Hawaii, Alaska, Singapore, North and East Africa Male>female, 40-65 years 470/million/yr (Hong Kong) Radiotherapy Abnormal 3* None identified EA-D (IgA) dominant
Difficult Carcinogens (salted fish)
*From two tumours
of virology, immunology and oncology, focused their attention on BL and the role played by EBV in the genesis of that tumour, more or less to the exclusion of NPC. Studies on the latter largely attracted epidemiologists and clinicians interested in discovering the genetic or lifestyle differences that might account for the restricted location and high incidence of NPC in certain parts of the world. It was obvious that factors other than infection with EBV, which is ubiquitous among the world's population, must be contributing to disease, and nitrosamines in salted fish, a staple diet in high incidence areas, received much attention13 as a 'cofactor' in disease, but the relationship remains, to date, unproved. The last few years have, however, seen an awakening interest in NPC, particularly among virologists and molecular biologists, and a greater exchange of information and materials between Eastern and Western parts of the world is greatly contributing to increased knowledge. The data presented in Table 1 also suggest that if EBV is a principal contributor to both BL and NPC, there must either exist different pathological strains of the virus (as in the case of herpes simplex, where two distinct types, HSV-1 and HSV-2, exist) or else the host cells, themselves, must be differentially permissive to viral gene expression, exerting their own control in the symbiotic relationship with the virus. With regard to the former, whereas there are variations among viral isolates, there are no convincing data relating to pathologically variant strains. Until recently there have been no data relevant to the latter. Our own interest in NPC began in the early 1980s when, with molecular techniques, we dissected the viral DNA into large overlapping fragments and used these to ascertain whether functions could be identified that altered the growth properties in culture of epithelial cells. One, among eight EBV DNA fragments tested in this way, had unusual properties. That is, in a mixture of primary primate cells (human or monkey) one specific fragment among them consistently stimulated epithelial cells in the population to proliferate indefinitely in culture and some of these became 'immortalized', whereas fibroblasts in the same population were unaffected in their growth'4 5. It was notable that this fragment in question contained none of the viral genes [the family of EBV nuclear antigens (EBNAs) or a membrane protein, designated latent membrane
protein (LMP)] that had hitherto been associated with viral stimulation and immortalization of Blymphocytes. Pagano and collaborators'6 analysed EBV transcriptional expression in several biopsies from American NPC patients and found messenger RNA from limited, but not wholly consistent, regions of the viral genome. Thus, the data at this stage pointed to cell specific EBV genes being expressed. This notion was considerably strengthened when we, with T Tursz and co-workers in Paris, began to analyse viral gene expression in detail in an NPC tumour from North Africa that had been established to grow in nude mice. Our data17 showed that the region we had already been studying with regard to cell growth stimulation was highly expressed, whereas other regions of the viral genome were essentially silent. Confirmatory evidence for these findings are now coming out of other laboratories, including one in Hong Kong focusing on EB viral gene expression in Asian NPCs (H-L Chen, M Lung and M H Ng, personal communication). The data being accumulated are consistent, not only with the notion that where EBV is concerned, the North African and Asian tumours appear to be similar, but virally important genes in the epithelial cell environment may not be totally identical to those related to the B-cell disease, BL. With the increasing attention now being given to NPC, we should soon know considerably more about the viral involvement in this disease. (Ironically, this change in the focus of interest has resulted, in some circles at least, in questions being raised regarding any role for EBV in BL. This concern is strengthened by the fact that there seems to be no involvement of EBV in non-endemic BL.) It has been known for some time that prior to the appearance of disease, IgA titres to EBV genes, such as the viral capsid antigen, VCA, rise sharply. In a wide-ranging serological screening of Chinese in NPC high incidence areas, people at risk of developing the disease were identified; follow-up studies of these showed that high IgA/VCA titres preceded clinical symptoms by 6-41 months'8. More recently, 130 asymptomatic patients from 6054 individuals with high EBV antibody titres were randomly recruited and examined by fibreoptic endoscopy'9. From among these, seven cases of NPC were detected. The tumour was confirmed as largely confined to the pharyngeal recess. These studies support the involvement of EBV
Journal of the Royal Society of Medicine Volume 85 February 1992
in the disease and at the same time point to the importance of the development of simple screening procedures for identifying people either carrying, or at risk of developing, this cancer considerably before it becomes evident by conventional means. Early patient management, as in most other human cancers, affords a much higher treatment success rate. Prognostic methods are available for NPC but, as we learn more about EBV gene expression in epithelial cells, can undoubtedly be improved. Dr George Klein (Department of Tumor Biology, Karolinska Institutet, Stockholm, Sweden) spoke on 'The Relative Role of the EBV Transformation and the Ig/myc Translocation in the Genesis of Burkitt's Lymphoma'. Translocation of the c-myc protooncogene to an immunoglobulin locus is a rate-limiting step in the genesis of three B-cell derived tumours: Burkitt lymphoma (BL) in humans, mouse plasmacytoma (MPG) and rat immunocytoma (RIO)20. The translocation must represent a rate-limiting step in the genesis of all three tumours. Its consequences have best been analysed in BL. They involve a nonimmunological and an immunological component. The former acts by preventing the B-cell from leaving the cycling compartment and entering the resting stage when programmed to do so. The latter acts by the down-regulation of certain HLA class I polymorphic specificities, adhesion molecules and EBV-encoded proteins. According to Dr Klein's interpretation, the translocation fixes the BL cell in a phenotypic window that can be referred to as 'a resting cell that is not resting'. The linking of c-myc to immunoglobulin sequences in a B-cell leads to constitutive myc expression. In spite of its switch to a 'resting phonotype', the cell is unable to leave the cycling compartment2l. Normally, EBV-carrying B-blasts would face immune controls. They can only proliferate in immunodefectives, as in XLP or transplant lymphoma. Due to its 'resting' rather than B-blast phenotype and the correlated defective expression of certain EBV, HLA and adhesion molecules, however, the BL cell is not controlled by the immune response. The down regulation of EBNA 2-6 and latent membrane protein (LMP) in the BL cell were also considered in relation to viral latency. The exclusive expression of EBNA 1 in the BL cell may reflect the adaptation of the virus to prolonged persistence in long lived resting B-cells.
'Hepatitis B viruses and hepatocellular carcinoma' was the topic presented by Dr Marie Annick Buendia (Unite de Recombinaison et Expression Genetique (Inserm U163), Institut Pasteur, Paris, France). Primary hepatocellular carcinoma (HCC), one ofthe most common cancers in many parts of the world, also shows seroepidemiological association with a viral infection: hepatitis B virus (HBV) has been clearly identified as a major cause of HOC. During the last decade, extensive studies, mainly focused on the HBV life cycle, the function of HBV proteins, the regulation of viral gene expression and the mechanisms of viral DNA integration into host cell DNA, have sustained different hypotheses on the viral contribution to carcinogenesis. Arguments favouring both direct and indirect mechanisms have accumulated recently. Integrated HBV sequences,
found in most HCCs examined, might play a part in carcinogenesis, either directly, through insertional activation of cellular genes, or indirectly, by altering the cell growth control in trans. In two independent tumours, HBV insertion sites have been identified as the receptor B for retinoic acid and the cyclin A genes22'23. In other tumours, truncated products of integrated viral X gene and PreS2/S region have been shown to harbour transcriptional trans-activation properties24'25. More indirect mechanisms, involving prolonged liver cell injury mediated by the antiviral immune response, and continuous hepatocyte regeneration, have also been proposed. Accordingly, overexpression of unsecretable, large HBV envelope protein in transgenic mice has been correlated with liver necrosis, inflammation and eventually HCC26. Much less attention, however, has been paid to the underlying cellular mechanisms leading to malignant transformation in response to viral induction. To date, no unifying model, accounting for the contribution of viral and cellular factors to liver oncogenesis, has been proposed. In contrast, studies of woodchuck liver tumours associated with chronic infection with woodchuck hepatitis virus (WHV) have revealed a more simple pattern. WHV, a virus closely related to HBV, is used as a model for liver disease and hepatocarcinogenesis in humans. More than 80% of carrier animals succumb to HCC after 4 years; moreover, the oncogenic properties of WHV, in the absence of carcinogenic cofactors, have been demonstrated by the induction of HCC in experimentally infected, newborn woodchucks. In a recent analysis of 30 woodchuck liver tumours, we have observed frequent activation of myc family genes, either by insertion of WHV DNA or genetic rearrangement, or by a trans-acting mechanism. In 30% of the tumours analysed, viral integration occurred within either c-myc, N-myc or a newly identified N-myc retroposon27 25. We have shown that comparable mechanisms, leading to deregulated expression of myc genes, operate in the development of tumours induced by hepatitis virus or by nonacute retroviruses in rodents. The contribution of the mutated myc genes to hepatocyte transformation has been demonstrated by the induction of primary liver tumours in transgenic mice carrying a woodchuck myc gene activated by hepadnavirus insertion. These results show that insertional activation of cellular proto-oncogenes by WHV DNA is a common event in woodchuck HCC, and implicate myc family genes in the genesis of most of these tumours.
Professor Howard Thomas (Department of Medicine, St Mary's Hospital, Medical School) also spoke about the association of HBV with HCC. Dr James Neil (Beatson Institute for 'Cancer Research, Glasgow) introduced the session on retroviruses, with his talk on 'Molecular pathogenesis of feline leukaemia vireus. Genetic damage arising from proviral insertion and recombination is now familiar as an explanation of the oncogeneity of type-C retroviruses. Retroviral 'tagging' of host cell genes by these processes underlies their pre-eminent role in elucidating the genetic basis of cancer. Feline leukaemia virus (FeLV) has been one ofthree highly productive models in this field. However, in natural FeLV infection, tumours
117
118
Journal of the Royal Society of Medicine Volume 85 February 1992
develop only rarely, and after a long latent period. To understand the basis of oncogenesis by these viral agents clearly requires a deeper knowledge of the virus-host interaction. Although the infectious, horizontally spread form of FeLV (FeLV-A) is remarkably conserved, its replication in vivo generates diverse variants which are associated with specific acute diseases. Thus, acutely pathogenic isolates of FeLV are invariably mixed populations of the replication-competent FeLVA and one or more variant genome. The reproduction of acute disease requires both the common and variant genomes to be present. Although FeLV variants are important mediators of acute disease they may thereby hamper virus dissemination and are not themselves efficiently transmitted under natural
conditions29. The variants include envelope gene mutants (FeLVC, FeLV-FAIDS) and recombinants with endogenous proviruses (FeLV-B) some of which have expanded host range and increased pathogenic potential30. Another site at which FeLV evolves from the prototype is in the proviral long terminal repeats where duplications of transcriptional enhancer elements are often present in tumour derived proviruses31. A third class of FeLV variant is represented by oncogene containing recombinants. However, acutely oncogenic variants appear to arise only rarely since neoplastic disease is an uncommon outcome of natural virus infection. Our recent work has focused mainly on the aetiology of the most frequently observed FeLV-associated neoplasm, thymic lymphosarcoma. Many of these tumours involve activation of the c-myc gene, either by proviral insertion or transduction. However, even where potent FeLV v-myc oncogenes are present, the tumours are of clonal origin and may have an immortal, factor-independent growth phenotype which has not so far been reproduced by in vitro infection32. It seems, therefore that other oncogenic events mediate progression of tumours initiated by an activated c-myc allele. By analysing proviral integration sites in tumours induced by FeLV-myc recombinant viruses, we have recently identified two common regions of
proviral insertion, designated fit-1 and fit-2. Gene fragments from these loci, which have no obvious homologue in gene sequence data-banks, detect two proviral integration clusters in four out of 13 independent tumours (H Tsujimoto and J C Neil, unpublished results). In one tumour (T17) a second transductant was present in addition to the myccontaining FeLV in the form of multiple proviruses carrying the complete coding sequence of a T-cell antigen receptor ,3-chain gene33. Although experimental infection with this recombinant has failed to demonstrate oncogenic potential for secondary hosts, an antigen or MHC-restricted oncogenic function remains plausible although difficult to evaluate in the outbred cat. Our current working model for the induction of thymic lymphosarcoma by FeLV is that accidents of virus replication disrupt the sensitive process of thymic education, whereby potentially autoreactive T-lymphocytes are normally recognized and induced to undergo programmed cell death. To test this as a general hypothesis in a more defined system, we have produced transgenic mouse lines in which the oncogenicity of c-myc is directed to the T-cell lineage by linking it to a dominant control region from the
human CD2 gene. These mice develop sporadic thymic tumours which can be greatly accelerated by neonatal infection with Moloney murine leukaemia virus (M Stewart, E Cameron, D Onions and J C Neil, unpublished). We believe that this will be an invaluable system in which to test the roles of the T-cell receptor repertoire and of loci identified by proviral insertion. Professor Jonathan Weber (Department of Genitourinary Medicine and Communicable Diseases, St Mary's Hospital) spoke on 'Human T-cell lymphotropic viruses'. Dr Valerie Beral (ICRF Cancer Epidemiology Unit, Oxford) spoke on 'HIV, AIDS and cancer'. In Europe and the USA one in five people with AIDS also have cancer, Kaposi's sarcoma and non-Hodgkin's lymphoma are the most frequent malignancies which occur in HIV-infected individuals. These cancers are rare in H1V uninfected people, except in other immunosuppressed groups such as transplant recipients. Thus the HIV-associated immunosuppression plays a major role in the development of these tumours. It is widely believed that the cancers which occur in the immunosuppressed have a viral aetiology. Kaposi's sarcoma is nevertheless 300 times as common in AIDS patients than in other immunosuppressed groups. It has been reported in 20% of all patients with AIDS in Britain, and is almost entirely confined to men who acquired the HIV by homosexual contact and, to a lesser extent, to heterosexuals who acquired their HIV infection in Africa3'435. Kaposi's sarcoma is, however, rare in patients with AIDS who acquired the HIV by non-sexual routes. The evidence to hand suggests that Kaposi's sarcoma in people with AIDS is caused by an as yet unidentified infectious agent and that the immunosuppression associated with HIV infection leads to clinical Kaposi's sarcoma in an otherwise asymptomatic carrier of the agent. That the risk of Kaposi's sarcoma is much higher in those who acquired their HIV sexually than by other means suggests that a major mode of transmission of the agent is sexual. Furthermore the higher risk of Kaposi's sarcoma in homosexuals than heterosexuals suggests that some aspect of homosexual behaviour facilitates the transmission of the agent. Non-Hodgkin's lymphoma occurs in about 3% of subjects with AIDS, irrespective of their means of acquiring HIV infection36. The lymphomas are clinically atypical, often arising in extrandodal sites. Primary lymphoma of the brain, which is normally exceedingly rare, accounts for about one fifth of the lymphomas which develop in patients with AIDS. The lymphomas tend to be high-grade lesions, most being classified as immunoblastic lymphomas. Burkitt's lymphoma also occurs in patients with AIDS and this is surprising, since Burkitt's lymphomas do not arise in other immunosuppressed individuals, whereas immunoblastic lymphomas do. Apart from the role of immunosuppression, the reason why lymphomas develop in people with AIDS is unclear. Infectious agents are believed to be involved and EBV may well play a part. EBV is commonly associated with African Burkitt's lymphoma not in other forms of lymphomas which arise in the immunosuppressed. Its association with HIV-associated lymphomas is not so clear. Whether the risk of other cancers is increased in HIV-infected individuals is still unknown.
Journal of the Royal Society of Medicine Volume 85 February 1992
Likely candidates are cancers with a known or suspected viral origin, such as hepatoma, cervical cancer and Hodgkin's disease, since the HIV-induced immunosuppression may well accelerate the process of viral carcinogenesis.
Jennifer M Best Department of Virology St Thomas' Hospital Lambeth Palace Road London SEl 7EH References 1 zur Hausen H. Papillomaviruses in anogenital cancer as a model to understanding the role of viruses in human cancers. Cancer Res 1989;49:4677-81 2 Jochmus-Kudielka I, Schneider A, Braun R, et al. Antibodies against the human papillomavirus type 16 early proteins in human sera: correlation of anti-E7 reactivity with cervical cancer. J Natl Cancer Inst 1989;81:1698-1704 3 Werness BA, Levine AJ, Howley PM. Association of human papillomavirus types 16 and 18 E6 proteins with p53. Science 1990;248:76-9 4 Kinlen LJ, Sheil AGR, Peto J, Doll R. Collaborative United Kingdom Australasian study of cancer in patients treated with immunosuppressive drugs. BMJ 1979;ii:1461-6 5 Orth G. Epidermodysplasia verruciformis. In: Salzman NP, Howley PM, eds. The Papovaviridae. The papillomaviruses. New York: Plenum, 199-243 6 Rudlinger R, Smith IW, Bunney MH, Hunter JAA. Human papilloma virus infections in a group of renal transplant recipients. Br J Dermatol 1986;115:681-92 7 Barr BBB, Benton EC, McLaren K, et al. Human papilloma virus infection and skin cancer in renal allograft recipients. Lancet 1989;i:124-9 8 Steger G, Olszewsku M, Stockfleth E, Pfister H. Prevalence of antibodies to human papillomavirus type 8 in human sera. J Virol 1990;64:4399-406 9 Cancer statistics: registration, England and Wales 1984. HMSO, London 1988; Scottish Health Statistics 1986/7; Statistics and Research Bulletin, SR1/87 DHSS Belfast 1988. (Taken from report, Cancer Research Campaign, 1989) 10 Zeng Y, Ou B. Etiology and pathogenesis of nasopharyngeal carcinoma. China: The People's Publishing House, 1985 11 de-The G. Role of Epstein-Barr virus in human diseases: Infectious mononucleosis, Burkitt's lymphoma, and nasopharyngeal carcinoma. In: Klein G, ed. Viral oncology. New York: Raven Press, 1980:769-97 12 Old LJ, Boyse EA, Oettegen HF, et al. Precipitation antibody in human serum to an antigen present in cultured Burkitt's lymphoma cells. Proc Natl Acad Sci USA 1966;56:1699-704 13 Huang DP, Ho JHC, Saw D, Teoh TB. Carcinoma ofthe nasal and paranasal regions in rats fed Cantonese salted marine fish. In: de-The G, Ito Y, eds. Nasopharyngeal carcinoma; etiology and Control. Lyon: IARC, 1978:315-18 14 Griffin B, Karran L. Immortalization of monkey epithelial cells by specific fragments of Epstein-Barr virus DNA. Nature 1984;309:78-82 15 Karran L, Teo CG, King D, et aL Establishment of immortalized primate epithelial cells with sub-genomic EBV DNA. Int J Cancer 1990;46:763-72 16 Raab-Traub N, Hood R, Yang C-S, Henry B, Pagano JS. Epstein-Barr virus transcription in nasopharyngeal carcinoma. J Virol 1983;48:580-90
17 Hitt MM, Allday MJ, Hara T, Karran L, Jones MD, Bussons P, Tursz T, Ernberg I, Griffin BE. EBV gene expression in an NPC-related tumour. EMBO J 1989; 8:2639-51 18 Zeng Y, Shen SJ, Deng H, et al. Early nasopharyngeal carcinoma among IgA/VCA antibody positive individuals detected by anti-complement immunoenzymatic medicine. Chinese Med J 1984;97:155-7 19 Sham JST, Wei W, Zong Y-S, et al. Detection of subclinical nasopharyngeal carcinoma by fibreoptic endoscopy and multiple biopsy. Lancet 1990;335:371-3 20 Klein G. Viral latency and transformation: The strategy of Epstein-Barr virus. Cell 1989;58:5-8 21 Klein G. Multiple phenotypic consequences of the Ig/myc translocation in B-cell-derived tumors. Genes, Chromosomes Cancer 1989;1:3-8 22 Dejean A, Bougueleret L, Crzeschik KH, Tiollais P. Hepatitis B virus DNA integration in a sequence homologous to v-erb-A and steroid receptor genes in a hepatocellular carcinoma. Nature 1986;322:70-2 23 Wang J, Chenivesse X, Henglein B, Brechot C. Hepatitis B virus integration in a cyclin A gene in a hepatocellular carcinoma. Nature 1990;343:555-7 24 Wollersheim M, Debelka U, Hofschneider PH. A transactivating function encoded in the hepatitis B virus X gene is conserved in the integrated state. Oncogene 1988;3:545-54 25 Kekule AS, Lauer U, Meyer M, Caselman WH, Hofschneider PH, Koshy R. The pre S2/S region of integrated hepatitis B virus DNA encodes a transcriptional transactivator. Nature 1990;342:457-61 26 Chisari FV, Klopchin K, Moriyama T, et aL Molecular pathogenesis of hepatocellular carcinoma in hepatitis B virus transgenic mice. Cell 1989;59:1145-56 27 Hsu TY, Moroy T, Etiemble J, et aL Activation of c-myc by Woodchuck hepatitis virus insertion in hepatocellular carcinoma. Cell 1988;55:627-35 28 Fourel G, Trepo C, Bougueleret L, et al. Frequent activation of N-myc genes of hepadnavirus insertion in woodchuck liver tumours. Nature 1990;347:294-8 29 Neil JC, Fulton R, Rigby M, Stewart M. Evolution of pathogenic and oncogenic variants of feline leukaemia virus. Curr Top Microbiol Immunol 1991, in press 30 Stewart MA, Warnock M, Wheeler A, et al. Nucleotide sequences of a feline leukemia virus subgroup A envelope gene and long terminal repeat and evidence for the recombinational origin of subgroup B viruses. J Virol 1986;58:825-34 31 Fulton R, Plumb M, Shield L, Neil JC. Structural diversity and nuclear protein binding sites in the long terminal repeats of feline leukaemia virus. J Virol 1990;64:1675-82 32 Onions D, Lees G, Forrest D, Neil JC. Recombinant feline viruses containing the myc gene rapidly produce clonal tumours expressing T-cell antigen receptor gene transcripts. Int J Cancer 1987;40:40-5 33 Fulton R, Forrest D, McFarlane R, Onions D, Neil JC. Retroviral transduction of T-cell antigen receptor Bchain and myc genes. Nature 1987;326:190-4 34 Beral V, Peterman TA, Berkelman R, Jaffe HW. Kaposi's sarcoma among persons with AIDS: a sexually transmitted infection? Lancet 1990;335:123-8 35 Beral V, Bull D, Jaffe H, Evans B, Gill N, Tillett H, Swerdlow AJ. Is risk of Kaposi's sarcoma in AIDS patients in Britain increased if sexual partners come from United States or Africa? BMJ 1991;302:624-5 36 Beral V, Peterman T, Berkelman R, Jaffe H. AIDSassociated non-Hodgkin's lymphoma. Lancet 1991; 337:805-9
119
