Personal view
Endemic Burkitt’s lymphoma
Is endemic Burkitt’s lymphoma an alliance between three infections and a tumour promoter? Malaria and Epstein-Barr virus (EBV), recognised cofactors for endemic Burkitt’s lymphoma, are ubiquitous within the lymphoma belt of Africa, and, unless other cofactors are involved, the tumour should be much more common than it is. Malaria and EBV alone cannot account for the occasional shifting foci and space-time case clusters of endemic Burkitt’s lymphoma. Arboviruses and plant tumour promoters are other possible local cofactors that could explain such characteristics. The geographical and age distributions of endemic Burkitt’s lymphoma parallel those of potentially oncogenic, mosquito-borne arboviruses. Arboviruses seem to be associated with case clusters of endemic Burkitt’s lymphoma, and symptoms compatible with arbovirus infection have been seen immediately before the onset of the tumour. RNA and DNA viruses, including EBV, are promoted by extracts of a commonly used plant, Euphorbia tirucalli, the distribution of which coincides with the boundaries of the lymphoma belt. Extracts of E tirucalli are tumour promoters and can induce the characteristic 8;14 translocation of endemic Burkitt’s lymphoma in EBV-infected cell-lines. They also activate latent EBV in infected cells, enhance EBVmediated cell transformation, and modulate EBV-specific immunity. Lancet Oncol 2004; 5: 738–46
Burkitt’s lymphoma (figure 1) has been called the Rosetta stone of oncology and is a model of human cancer.1 It is a high-grade non-Hodgkin lymphoma that occurs sporadically worldwide, but is endemic in Papua New Guinea and in the lymphoma belt of Africa (figure 2). This belt stretches from about 10º north to 10º south of the equator. Within these latitudes, the lymphoma is found in regions where the mean minimum temperature exceeds 15·5ºC and the rainfall is higher than 50 mL per year.2 It is more common in damp, humid regions, and is absent in arid regions. Within the lymphoma belt, endemic Burkitt’s lymphoma accounts for up to 74% of childhood malignant disorders. Incidence rates of Burkitt’s lymphoma vary from low in industrialised countries such as the USA and in western Europe, through intermediate rates in countries such as Algeria, up to high rates such a 4–10 per 100 000 in children aged younger than 15 years who live in the lymphoma belt.3 Rates of Burkitt’s lymphoma have increased in low-incidence countries since the 1980s, preceding the advent of HIV/AIDS, which has also contributed to the rise in non-Hodgkin lymphomas.4 Cases
738
Courtesy of R Nitzsche
C A van den Bosch
Figure 1. African child with facial lesions of Burkitt's lymphoma.
of Burkitt’s lymphoma peaked in the late 1980s in lowincidence countries, whereas in the lymphoma belt, they have tripled and continue to increase.5 Endemic Burkitt’s lymphoma most commonly presents as a facial tumour (figure 1), but it can also present in other ways, such as abdominal tumours (figure 3). In sporadic Burkitt’s lymphoma, including countries with intermediate rates, abdominal tumours are commonest. In addition to the two main forms of Burkitt’s lymphoma, HIVassociated Burkitt’s lymphoma has been recognised, which tends to present similarly to sporadic Burkitt’s lymphoma outside the lymphoma belt. Within the lymphoma belt, it occurs mainly in adults, presents atypically, and accounts for up to two-thirds of Burkitt’s lymphoma.5 CAvdB is Consultant in Communicable Disease Control, Surrey Health Protection Unit, Health Protection Agency and Honorary Clinical Lecturer at the Wolfson Institute of Preventive Medicine, Barts and the London Queen Mary’s School of Medicine and Dentistry, London, UK. Correspondence: Dr C van den Bosch, Surrey Health Protection Unit, Century House, 26 Bridge Street, Leatherhead, KT22 8BZ UK. Tel: + 41 (0)1372 869 625. Fax: +41 (0)1372 869884. Email:
[email protected]
Oncology Vol 5 December 2004
http://oncology.thelancet.com
Personal view
Endemic Burkitt’s lymphoma
Figure 2. The lymphoma belt: Burkitt’s lymphoma is endemic in regions with mean minimum temperatures that exceed 15·5ºC and yearly rainfall of higher than 50 mL, which stretch from about 10º north to 10º south of the equator.
A combination of malaria and EBV seems to boost the incidence of endemic Burkitt’s lymphoma in the lymphoma belt by a factor of 100–150.1 The EBV genome is incorporated into 96% of endemic Burkitt’s lymphomas, but in only 10–15% of sporadic cases in France and in 80% in Algeria.1,6 A gradient exists for the tumour subtype, the number of patients who have EBV positivity and the type of translocation, strongly suggesting that other environmental factors are involved.7 The presence of EBV in Burkitt’s lymphoma probably reflects the prevailing socioeconomic status and age at seroconversion of each population of patients. Despite the HIV epidemic in sub-Saharan Africa, EBV is still present in 94% of cases of endemic Burkitt’s lymphoma in children aged younger than 15 years in whom it is not associated with HIV.8 HIV has been associated with some cases of sporadic Burkitt’s lymphoma: HIV-associated Burkitt’s lymphoma has increased worldwide, where 30–50% of cases are EBV-positive,9 and within the lymphoma belt, where it has increased three-fold, especially in adults, with atypical presentations,5 but mostly EBVpositive tumours, attributable to the HIV epidemic.8 Burkitt’s lymphoma is a B-cell lymphoma with a high replication rate. Almost all the cells are in cycle and have a doubling time of 24–48 h,10 which is consistent with the rapid clinical progression of the tumour. Cells in both endemic and sporadic forms of Burkitt’s lymphoma are B lymphocytes that have rearranged Ig genes and contain one of three translocations6 involving the long arm of chromosome 8, of which 8;14 is the commonest. The translocation occurs at the site of the MYC oncogene on chromosome 8 and the Ig heavy-chain genes on chromosome 14. Less common translocations occur between chromosome 8 and chromosome 2 (site of light-chain
Oncology Vol 5 December 2004
gene) or chromosome 22 (site of light-chain gene). Endemic and sporadic forms of Burkitt’s lymphoma have different breakpoints within both the MYC locus on chromosome 8 and the Ig heavy-chain locus on chromosome 14.7 EBV1 and holoendemic malaria1,6,7,11 are recognised as important cofactors of endemic Burkitt’s lymphoma, yet this disease can occur in the absence of both of these infections. Both infections are ubiquitous in the lymphoma belt and, if early EBV infection and malaria were the only prerequisites for endemic Burkitt’s lymphoma, then the tumour should be much commoner than it is. In this Personal view, I discuss the epidemiological evidence for the pathogenesis of endemic Burkitt’s lymphoma and propose that malaria and early EBV infection precede the translocation that occurs as the result of an interaction between EBV and a plant tumour promoter, and that these final stages of tumourigenesis are enhanced by acute arboviral infection. The pathogenesis of sporadic and HIVassociated Burkitt’s lymphoma are not discussed because the geographical gradient for the clinical, molecular, and cytological differences and the varying degrees of EBV positivity and different chromosomal breakpoint sites between the different types of Burkitt’s lymphoma suggest that, outside the lymphoma belt, other, unrecognised, environmental factors are important in the pathogenesis of this tumour, and these factors probably differ throughout the world.
Pathogenesis Klein12 proposed a multistep hypothesis for the pathogenesis of endemic Burkitt’s lymphoma with early, heavy infection with EBV in children as the first stage, resulting in the immortalisation of B lymphocytes and some immune tolerance, permitting proliferation of infected cells. Heavy malarial infection would stimulate expansion of the B-cell pool and suppress T cells involved in EBV control. The final stage of tumorigenesis would be the development of the translocation leading to deregulation of MYC and the development of a malignant clone.
Figure 3. Abdominal and facial tumours from Burkitt’s lymphoma.
http://oncology.thelancet.com
739
Personal view Bornkamm and colleagues13 thought that the Burkitt’s lymphoma translocation occurred when B cells were rearranging their chromosomes, as suggested by the constant involvement of one of the Ig loci. Intense immunological stimulation from holoendemic malaria would give rise to a large pool of B cells, which would increase the risk of developing translocations. They argued that once the translocation involving MYC had arisen, the third step would be infection and immortalisation by EBV of a cell that had already rearranged its Ig genes. Klein’s theory seems more probable than that of Bornkamm and colleagues because EBV infection occurs in infants in Africa.1 The rapid growth rate of Burkitt’s lymphoma cells could be an immediate consequence of the translocation, which juxtaposes a gene involved in IgG production with MYC—an oncogene involved in cellular proliferation—which becomes deregulated. Arboviral infection was proposed as a cofactor by Haddow in 1964.14 Some arboviruses have been shown to have oncogenic properties,15,16 and some plant extracts can act as tumour promoters.17 Immunotypic data and molecular evidence suggest that the translocation occurs in a B cell that is actively arranging its Ig genes, and that some differentiation of the cell then occurs as it acquires the characteristics of malignant disease. I suggest that the rearrangement of the Ig genes18 involved in the Burkitt’s lymphoma translocation takes place during the arboviral infection and that the action of the two potentially oncogenic viruses—ie, the arbovirus and EBV—are potentiated by a plant tumour promoter. Epidemics of arboviruses could explain the occurrence of space-time case-clusters of endemic Burkitt’s lymphoma.3
Role of EBV EBV is a DNA herpesvirus that occurs worldwide, infecting more than 90% of the population. Infection is mostly asymptomatic when it occurs at an early age, but, if the age of onset is delayed, it gives rise to infectious mononucleosis. Infection in Africa occurs at a very young age,1 and consequently, immune tolerance to the virus usually develops. EBV infects and immortalises B lymphocytes in vivo and in vitro, resulting in polyclonal activation and proliferation, which are usually controlled by inhibitory immunological mechanisms, with EBV-specific cytotoxic T cells playing a crucial part.19 Primary EBV infection is usually followed by latency, but in immunodeficiency, the proliferation of cells can proceed unchecked, and cells sometimes evolve from a polyclonal reactive process to a monoclonal malignant lymphoma.20 In fatal infectious mononucleosis, moreover, multiple distinct clones, some of which involve MYC rearrangement, have been detected within a few days of primary EBV infection.21 EBV could act as a cofactor by increasing the size of the B-cell pool and by transforming lymphocytes, thereby increasing the chances of a translocation occurring, or it could have a more direct role in tumorigenesis, acting together with the changes induced by the MYC translocation.6,7,18 Endemic Burkitt’s lymphoma seems to be associated with raised concentrations of some EBV
740
Endemic Burkitt’s lymphoma
antibodies, suggesting that EBV is not merely a passenger in this disease. Compared with matched controls, patients with endemic Burkitt’s lymphoma were shown to have significantly higher titres of IgG antibody to EBV viral capsid antigen (VCA) up to 6 years before the onset of the lymphoma.1 However, the VCA antibody titres did not rise after the onset of endemic Burkitt’s lymphoma, suggesting that chronic rather than acute infection with EBV was relevant to tumorigenesis. No other EBV antibody showed significant differences between pre-endemic Burkitt’s lymphoma cases and controls. EBV early antigen concentrations rise as the disease develops, and decline after treatment.22 The amount of cellular cytotoxicity mediated by EBV-specific antibodies seems to be important in the outlook for patients with endemic Burkitt’s lymphoma, further confirming the relation between this disease and EBV.23 EBV can exist in several different types of latency. Latency 1 has been identified in biopsy samples from patients with Burkitt’s lymphoma and in early-passage Burkitt’s lymphoma lines, in which many EBV-encoded RNA (EBER) transcripts are found, and EBNA1 is usually expressed in the absence of other EBV proteins.24 EBNA1, an EBV antigen that is indispensable for B-cell transformation, seems to maintain the EBV episome in latent infected cells.25 EBNA1 can bind both RNA and DNA, which could affect the expression of viral or cellular genes,26 upregulate the recombinase activating genes that mediate variable diversity joining (VDJ) recombination, which are usually expressed only in immature lymphoid cells,27 and can enhance B-cell immortalisation by several thousand times.28 In transgenic mice, EBNA1 and Myc, the mouse analogue of the human oncogene MYC, cooperate in the development of lymphomas, suggesting that the same process could occur in humans.29 However, EBNA1 does not seem to be oncogenic on its own, since it is consistently expressed in cells infected with EBV, including latently infected cells, and the EBER transcripts seem to be needed for a malignant phenotype and resistance to apoptosis.30 EBER1 and EBER2 are small, nonpolyadenylated nuclear RNAs. They upregulate BCL2, to inhibit apoptosis by binding to protein kinase, blocking the signalling of interferon ␣, and inducing colony growth of cells in agar.31 Furthermore, a binding site for MYC has been found in the promoter for EBER1, suggesting that it cooperates with MYC, and that EBV has a role in the development of lymphomas.31
Role of malaria Holoendemic malaria undoubtedly has a major immunomodulatory role in endemic Burkitt’s lymphoma in the lymphoma belt. It contributes to the greatly increased numbers of EBV-positive Burkitt’s lymphoma in this region and might facilitate the transformation of cells. A malarial mitogen32 that produces a polyclonal expansion of B lymphocytes, and, in acute attacks of malaria, the number of circulating EBV-positive lymphocytes is increased by about five times.33 Hyperendemic malaria also inhibits EBVspecific suppressor T cells, thus expanding the pool of EBV-positive lymphocytes in two ways. Spontaneous
Oncology Vol 5 December 2004
http://oncology.thelancet.com
Personal view
Endemic Burkitt’s lymphoma
transformation has also been noted in cell cultures from people living in hyperendemic malaria regions.34 Malarial infection in holoendemic regions is common from infancy throughout childhood, until some immunity and tolerance is acquired by the age of puberty.35 The reduced incidence of endemic Burkitt’s lymphoma in people with the sickle-cell trait36 and in regions where malaria has been eradicated,11 together with the delayed age of onset of endemic Burkitt’s lymphoma in immigrants from higher, malaria-free altitudes, to lower malarial regions,11 have been regarded as evidence in favour of malaria acting as a cofactor in Burkitt’s lymphoma. Malaria-eradication programmes in Italy have also been linked with reduction in the number of cases of another herpesvirus-associated malignant disease—Kaposi’s sarcoma37—showing that the tumour-suppressor effect of malaria eradication is not confined to endemic Burkitt’s lymphoma. However, malaria could be both an amplifier and a confounder in endemic Burkitt’s lymphoma because, apart from the inverse association with the sickle-cell trait, which is controversial, these associations could also be the result of effects of mosquito-borne arboviruses acting as cofactors. A malaria-prophylaxis programme in Tanzania coincided with a reduction in the number of cases of endemic Burkitt’s lymphoma, but the reduction actually antedated the malaria suppression programme and was also seen in the adjoining district where there had been no intervention, suggesting that other factors might be active. A lag period before the incidence of endemic Burkitt’s lymphoma returned to its former level was seen after parasitaemia levels returned to normal and suppression had been stopped.38 These findings cannot be explained by a reduction in malaria from prophylaxis, because the reduction antedated the programme, but they could be the result of cyclical changes in other cofactors. In holoendemic regions, malaria occurs more or less throughout the year and is seasonal, but not cyclical, in regions where transmission is less intense. Outbreaks attributed to malaria are confined to regions where malaria is not usually seen, or where malarial activity increases unusually as a result of ecological changes.35 Cyclical changes are also not a feature of EBV infection in Africa, but are typical of mosquito-borne arboviruses; epidemics with arboviruses such as chikungunya or yellow fever occur regularly at variable intervals of 3–20 years.
Age and acquisition of antibodies In the lymphoma belt, endemic Burkitt’s lymphoma is a disease of childhood with a peak around 5–8 years of age. No case has been recorded in children younger than 1 year and most patients are younger than 10 years of age. Burkitt’s lymphoma in children younger than 4 years or older than 20 years are rare in endemic regions. Endemic Burkitt’s lymphoma is seen in adults who have been brought up in highland regions when they move into lower areas. This age distribution does not accord with the pattern of acquisition of antibodies to EBV. Infection with EBV occurs at a young age in sub-Saharan Africa, usually within the first year of life,1 in regions conforming to the climatic requirements of the lymphoma belt, and in higher, colder, and dryer regions.
Oncology Vol 5 December 2004
Most children in tropical Africa have acquired antibodies to EBV by their second birthday. The age distribution of endemic Burkitt’s lymphoma is similar to the pattern of acquisition to immunity to malaria.34,35 Malaria in the lymphoma belt is rare in children younger than 3 months, although congenital malaria can occur and is common throughout childhood, with the main adverse effect on health being in children younger than 6 years. Some tolerance and immunity is established by 8–14 years of age, depending on the amount of malarial endemnicity. However, the age distribution of acquisition of antibodies to arboviruses mimics the age distribution of endemic Burkitt’s lymphoma almost exactly.14,39 Neither endemic Burkitt’s lymphoma nor arbovirus infections are seen in children younger than 2 years, these infections peak at the age of 5–8 years, and, in the absence of HIV, are rarely seen in individuals older than 20 years, except in those who originated from higher altitudes.11 The age distribution of endemic Burkitt’s lymphoma and rates of acquisition of antibodies to arboviruses coincide up to the age of 10 years, after which the fit is less good, but this finding could be attributed to the pattern of age estimation seen in the teenage years in rural African hospitals.39 The age distribution of endemic Burkitt’s lymphoma thus accords best with the pattern of acquisition of antibodies to arboviruses, rather than with EBV or malaria.39
Arboviruses and herbal remedies Arboviruses, with their cycles of activity every 3–20 years, could, if acting as cofactors, provide an explanation for the variation in incidence of endemic Burkitt’s lymphoma seen before, during, and after the Tanzanian malaria prophylaxis programme.38 Because mosquito-borne arboviruses and malaria share the same insect vector, both would be reduced by programmes of malarial or mosquito eradication, but arboviruses would not be affected by malarial prophylaxis programmes. Arboviruses as cofactors, with their cyclical activity, could equally explain, either wholly or in part, not only the reduction in the number of cases of endemic Burkitt’s lymphoma seen after malarial eradication programmes, but also the case clusters and shifting foci of cases of endemic Burkitt’s lymphoma within the lymphoma belt3 and the older age of onset of endemic Burkitt’s lymphoma in cases emigrating from highlands that are free of malaria and arboviruses.11 Both endemic Burkitt’s lymphoma and yellow fever, a flavivirus and arbovirus, share the same temperature-determined altitude barrier2 of 1524 m in eastern Africa, whereas malaria epidemics have been recorded at 2438 m. Arbovirus infections
Arboviruses are RNA viruses transmitted by insect vectors such as mosquitoes and are found in both temperate and tropical regions. Arboviruses belong to several different families such as alphaviruses, flaviviruses, and Bwamba viruses. Arbovirus epidemics occur regularly in Africa, causing febrile illnesses with varied symptoms, including rashes, joint pains, encephalitis, jaundice, haemorrhagic manifestations, and, more rarely, shock syndrome.
http://oncology.thelancet.com
741
Personal view
Endemic Burkitt’s lymphoma
epidemics. Signs and symptoms compatible with arbovirus infection, including rashes, joint pains and fever, were reported in patients with endemic Burkitt’s lymphoma shortly before the Sign or Patients Time before Burkitt’s Chikungunya Chikungunya onset of the lymphoma, both in Uganda47 and symptom (%) lymphoma† IgG, IgM in Malawi.44 88 of 108 proven cases of endemic seroconversion Burkitt’s lymphoma had signs and symptoms Rash 9 (8) 8 (2–14) 5 5 consistent with arboviral illness, 51 patients Sore eyes 16 (18) 19 (7–28) 10 9 Joint pains 32 (37) 14 (2–28) 16 7 had typical signs and symptoms, and only ten Mouth ulcers 14 (16) 13 (3–21) 10 8 patients could not recall any signs of symptoms Fever 27 (31) 16 (1–56) 19 9 in the 3 months preceding the onset of the Bleeding 14 (16) 19 (14–28) 6 3 tumour (see table).44 Early studies in Uganda48 *Data taken from ref 44. †Data are median (range) in days. showed that patients with endemic Burkitt’s lymphoma and their relatives were more likely to be seropositive for some arboviruses, Subclinical cases are common in the flavivirus yellow fever, including chikungunya virus, than were controls, but in in which the ratio of subclinical to clinical cases is Malawi, the number of patients seropositive for seven to one. chikungunya IgG was significantly greater than in agematched controls—both local and hospital. 44 The late onset Case clusters during an arboviral epidemic of seroconversion seen in some patients with a history of Chikungunya fever, an alphavirus40 and arbovirus, has cycles arbovirus-like signs and symptoms in the 2 months before of activity every 5–10 years. An increased number of cases of tumour manifestation could be attributed to the abnormal endemic Burkitt’s lymphoma, most of which were positive immune function described in endemic Burkitt’s for the arbovirus chikungunya, and a southerly drift of foci lymphoma,48 consistent with three patients remaining of cases, were seen in 1987–88 in Malawi,3 when an epidemic seropositive for chikungunya IgM and seronegative for IgG of chikungunya fever spread from north to south. Clusters of over many months.44 Acute infection with an arbovirus could cases of endemic Burkitt’s lymphoma, which were closer thus explain why the B cells are rearranging their genes when together in time and space than would occur from chance, the translocation in endemic Burkitt’s lymphoma occurs. Chikungunya, an alpha virus, and yellow fever, a were seen in children older than 7 years in Uganda41–43 in 1961–65 and 1972–73 and in Malawi in 1987–89.3 69% of the flavivirus, have the potential to be oncogenic; they can Malawian patients seen during the chikungunya fever transform mouse brain cells16 and are tumorigenic in nude epidemic, including some in clusters, who were tested for mice.15 Hepatitis C, a flavivirus, is independently associated antibodies were seropositive for chikungunya virus IgM or with hepatocellular carcinoma in humans, and can act IgG at high titres on admission, or seroconverted to IgG synergistically with hepatitis B, a DNA virus.49 The during their first admission for endemic Burkitt’s replication of hepatitis C has also been shown to be lymphoma, suggesting that they had recently acquired the enhanced in the presence of EBNA1.50 In endemic Burkitt’s infection.44 The case clusters described in Uganda in lymphoma, an arbovirus, many of which are flaviviruses, 1961–6541,43 occurred at times of o’nyong-nyong virus could act synergistically with the DNA virus EBV, and its activity,45 and in 1972–73 during times of Bwamba virus replication be enhanced by EBNA1. Several researchers51,52 have shown that hepatitis B, C, and G (another flavivirus), activity.46 Antibodies to several arboviruses were much more occur more commonly in patients with B-cell common in patients with endemic Burkitt’s lymphoma and lymphoproliferative disorders than in controls, and it has their families than in controls, but no one arbovirus been tentatively proposed that the possibility of these viruses predominated.10,45 This finding would fit with several having a role in development of lymphomas is stronger for arboviruses having the ability to act as a cofactors.10,45 The hepatitis C and G than for hepatitis B. Although neither fact that seroconversion occurred in Malawi44 during the first hepatitis C nor G are, by definition, arboviruses, both are admission for endemic Burkitt’s lymphoma could be because flaviviruses, like many important arboviruses, and, if a role the time between the arboviral infection and the onset of the in the development of lymphomas is confirmed, the tumour was very short, which is consistent with the reports argument for arboviruses acting as cofactors in endemic of typical arboviral symptoms with means of 8–19 days Burkitt’s lymphoma would be strengthened. The before the onset of the tumour (see table),44,47 the tumour’s epidemiology of hepatitis B does not resemble that of rapid growth rate,3 and the abnormal Ig response seen in endemic Burkitt’s lymphoma and the epidemiology of the endemic Burkitt’s lymphoma.44,48 other two viruses has not been fully elucidated. Chikungunya virus is commonly found in southeast Asia Signs and symptoms of arbovirus infection in regions where Burkitt’s lymphoma is found, but is not It is possible that the case clusters of endemic Burkitt’s endemic. The lesser incidence of Burkitt’s lymphoma in this lymphoma are the result of arboviruses acting as cofactors, region could possibly be accounted for by genetic, cultural, precipitating the final stages of the pathogenesis of the or socioeconomic differences, or could relate to differing tumour, and that they are seen only during arbovirus epidemiology of falciparum malaria or arboviruses, or a later Characteristic arboviral signs and symptoms in patients with endemic Burkitt’s lymphoma immediately before development of lymphoma*
742
Oncology Vol 5 December 2004
http://oncology.thelancet.com
Personal view
Endemic Burkitt’s lymphoma
tions in B cells,59 thus suggesting that these substances have a role in development of lymphomas associated with EBV and with endemic Burkitt’s lymphoma.54,55,57
Action of plant tumour-promoter on EBV in vitro Cell lines of human lymphocytes immortalised by EBV after protracted culture develop chromosomal abnormalities that have some concordance with those seen in B-cell neoplasms.60 A very small number of aneuploid lymphoblastoid cell lines are tumorigenic in immunocompromised animals.60 Although the characteristic 8;14 chromosomal translocation of Burkitt’s lymphoma has not been seen to arise spontaneously in lymphoblastoid cell lines, when an extract of E tirucalli was added, continuous mitoses ensued in the lymphoblastoid cell lines, with chromosomal abnormalities in 10% of the mitoses seen over the course of 1 year.61 About 10% of these abnormalities were translocations involving the long arm of chromosome 8 with activation of the oncogene MYC. Cells with these abnormalities were tumorigenic in nude mice.61 Thus, E tirucalli, which is commonly used in the lymphoma belt as a toy, hedge, and a herbal remedy for paediatric complaints, including signs and symptoms of arbovirus infections,54,55,57 can produce translocations of the type seen in endemic Burkitt’s lymphoma in human lymphocytes in vitro when they are infected with EBV, and might also be able to do so in vivo. Figure 4. The Euphorbia tirucalli plant is a commonly used herbal remedy in Africa.
age of infection with EBV, since Burkitt’s lymphoma is seen in people of all ethnic origins in the lymphoma belt in apparently representative proportions. Standards of living in rural Africa and India are much the same.
Chapin’s zones The geographic distribution of endemic Burkitt’s lymphoma in Africa coincides with one of Chapin’s zones which circumscribe a region where some insects and plants are located because of the environmental conditions which they require.53 A common plant within the lymphoma belt, which has a similar distribution, is Euphorbia tirucalli (figure 4).54
Use of tumour-promoting herbal remedies E tirucalli, a herbal remedy commonly used in Africa, grows significantly more frequently at the homes of patients with endemic Burkitt’s lymphoma than in those of people without this disease.54–56 It contains a substance closely related to the tumour-promoter TPA (12-O-tetradecanoylphorbol13-acetate), which is derived from another euphorbia, Croton tiglium. E tirucalli is used medicinally and as hedging, and its rubbery latex is played with by children.57 It produces EBV-promoting substances, present in plant parts, and secreted into the soil around the plant in active form.56 In addition to activating the latent EBV in infected cells, these substances can also enhance EBV-mediated cell transformation,56 modulate EBV-specific T-cell activity and cellular immunity,58,59 and induce chromosomal transloca-
Oncology Vol 5 December 2004
Effect of plant tumour promoter on arboviruses Tanaka62 showed the potentiation of a carcinogen by the application of TPA to a nude mouse during infection with the arbovirus West Nile virus. The synergy between virus and carcinogen has been shown in many other models, including the interaction between hepatitis B and the carcinogen aflatoxin.63 TPA can also enhance production of complete virions of several viruses—both DNA viruses like EBV64 and RNA viruses such as retroviruses or arboviruses—and can amplify the expression of oncogenes such as MYC, known to be implicated in the pathogenesis of Burkitt’s lymphoma.65 The extracts of E tirucalli, which are tumour promoters, could possibly act synergistically with arboviruses, which are potentially oncogenic,15,16 and, in addition to their clastogenic effect on chromosomes and predilection for chromosome 8,61 could potentiate any oncogenic activity of EBV and the arbovirus. Other plants used medicinally in Africa also contain substances that can activate EBV in latently infected cells64 and could possibly have at least some of the other properties of the extracts of E tirucalli. Chikungunya virus and probably other arboviruses could be cofactors in the pathogenesis of endemic Burkitt’s lymphoma, as suggested by Haddow,2,14,39 and a tumour promoter such as E tirucalli could be the final stage in the induction of endemic Burkitt’s lymphoma. This combination of cofactors could act with a short latent period, which would fit with both the history of arbovirallike illnesses immediately preceding the onset of the tumour seen in Uganda47 and Malawi,44 and with the rapid doubling time of the tumour—66 h in one study.10 Cloning of
http://oncology.thelancet.com
743
Personal view translocation breakpoints from several endemic cases has shown involvement of VDJ recombinase in the genesis of the translocations,18 suggesting that endemic Burkitt’s lymphoma arises in a B cell that is in the process of actively arranging its IgG genes. However, other evidence suggests that the pattern of deletions and insertions seen in Ig VH DJH mutations occurs as the result of an antigen-driven selection process, and that the MYC/Ig translocation takes place as a result of hypermutation in B cells entering or transiting germinal centres.66–68 Such changes could occur in response to arboviral infection. The germinal centres of individuals who are undergoing germinal-centre hyperstimulation, as in hyperendemic malarial regions, are rich in EBER-positive cells, which are usually absent from the germinal centre. Niller31 has postulated that when an EBER-positive cell undergoes a translocation associated with Burkitt’s lymphoma in the germinal centre, MYC helps to open and upregulate the antiapoptotic EBER transcription units, and the balance between apoptosis and anti-apoptosis is thus permanently shifted in favour of cell survival. The cell will then have time to accumulate additional tumorigenic mutations, if these are necessary, or if it is already a Burkitt’s lymphoma cell, can expand under an appropriate growth stimulus. This theory would also fit with the action of the plant tumour promoter to induce the translocation and promote the EBV, and an acute infection with an arbovirus, as suggested here, could stimulate B-cell expansion.
Possible mechanisms of viral oncogenesis A mechanism for arboviral oncogenesis could be virusinduced immunosuppression, but interaction or potentiation of the two viruses, EBV and the arbovirus, is more likely because the increase in Burkitt’s lymphoma in endemic areas is all related to EBV. The arbovirus might act Further reading 1 Burkitt DP. An alternative hypothesis to a vectored virus. In: Burkitt DP, Wright DH (eds). Burkitt’s lymphoma. Edinburgh: Churchill Livingstone, 1970: 211. 2 De The G. Epstein-Barr virus and Burkitt’s lymphoma worldwide: the casual relationship revisited. In: Lenoir G, O’Conor G, Olweny CLM (eds). A human cancer model, Burkitt’s lymphoma. Lyon: IARC Scientific Publication, 1985; 60: 165–75. 3 Haluska FG, Tsujimoto Y, Croce CM. The molecular genetics of non-Hodgkin’s lymphomas. In: Magrath IT (ed). The nonHodgkin’s lymphomas. Philadelphia: Edward Arnold, 1990: 97–102. 4 Ito Y. Vegetable activators of the viral genome and the causation of Burkitt’s lymphoma and nasopharyngeal carcinoma. In: Epstein MA, Achong, BG (eds). The Epstein-Barr Virus: recent advances. London: Heinnemann Medical Books, 1985: 209–34. 5 Monath TP, Flavividae: flaviviruses(yellow fever, dengue, dengue haemorrhagic fever, Japanese encephalitis, St Louis encephalitis, tick-borne encephalitis) In: Mandell GL, Bennett JE, Dolin R (eds). Principles and practice of infectious diseases. 4th edn. New York. Churchill Livingstone, 1995: 1465–74. 6 Morrow RH. Epidemiological evidence for the role of Falciparum malaria in the pathogenesis of Burkitt’s Lymphoma. In: Lenoir G, O’Conor G, Olweny CLM (eds). A human cancer model, Burkitt’s lymphoma. Eds. Lyons: IARC Scientific Publications No 60, 1985: 177–84. 7 Rickinson AB, Gregory CD. Burkitt’s lymphoma. Trans R Soc Trop Med Hyg 1988; 82: 657–59.
744
Endemic Burkitt’s lymphoma
Search strategy and selection criteria The initial hypothesis was derived from background reading of work on endemic Burkitt’s lymphoma from 1958 onwards, augmented by searches of MEDLINE Plus, from 1966 to 2004, and PubMed databases for primary references. Search terms used were “Burkitt’s lymphoma” or “endemic Burkitts lymphoma”, in combination with “viruses”, “oncogenic viruses”, “Epstein-Barr virus”, “alphaviruses”, and “flaviviruses”. Secondary searches were done after a review of the published work derived from primary searches, including related articles and search terms.
by driving cell growth and hypermutation, enhancing the MYC deregulation so that any genetic errors induced by EBV and the tumour promoter would be fixed before DNA repair could be affected, or by increasing the tendency to chromosomal breakages, as seen in EBV-positive cell lines cultured for long periods,60 or in some other way, such as the enhancement of hepatitis-C replication by EBNA1.50 A nuclear phase of yellow fever and other arboviruses has been reported,69 so that interaction between the two viruses and the host’s chromosomes could occur in the nucleus. Single-stranded RNA viruses such as alphaviruses have been suggested to function as mRNA and as catalysts, and flaviviruses can form recombinants.70 In the new era of mixand-match designer virology, alphaviruses71 and flaviviruses72 are being developed both as vaccine vectors in their own right and to enhance the effectiveness of DNA vaccines.73 If arboviruses have a nuclear phase, there is also the possibility that they have an action similar to that of retroviruses, where the viral genome is integrated into the genome at the nuclear stage and can act to produce insertional oncogenesis. Arboviruses could also provide the explanation for the retroviral-like sequences that have been described in patients with Burkitt’s lymphoma and in their healthy parents,74 postulated to be silent in the parents, but active in an oncogenic capacity in young people. Such insertions of oncogenic viruses, along with translocations, amplification of proto-oncogenes, and deletion of tumoursuppressor genes occur preferentially in certain fragile sites of genomic DNA that are prone to breakage.75
The burden of proof Because the case-clusters described in Uganda41,43 coincided with widespread o’nyongnyong virus,45 and localised Bwamba virus, activity,45,46 and the Malawian case-clusters occurred during a chikungunya epidemic,3 these arboviruses would seem to be likely candidates for cofactors acting at the final stage of endemic Burkitt’s lymphoma tumorigenesis. If serum samples are still available from patients who were part of the clusters or from the prospective study undertaken in Uganda,1 they could be tested for evidence of recent infection by a range of arboviruses at the onset of the lymphoma. If high titres of antibodies, or seroconversion early in the course of the tumour were detected in cases significantly more often than in controls, this finding would argue in favour of the arbovirus-cofactor hypothesis. Prospective case-control
Oncology Vol 5 December 2004
http://oncology.thelancet.com
Personal view
Endemic Burkitt’s lymphoma
studies are also needed. The arbovirus cofactor has consistency, temporality, and biological plausibility in its favour. If subsequent work elucidates the possible role for both arboviruses and plant extracts as final cofactors in the pathogenesis of endemic Burkitt’s lymphoma, which work in conjunction with EBV, and, to a lesser extent, with malaria, appropriate health education could be instituted. Measures advocated would include destruction of mosquito breeding sites, early treatment of malaria, early recognition and treatment of endemic Burkitt’s lymphoma and advice to traditional healers on the dangers of certain remedies. The new EBV vaccine, if deployed widely in the lymphoma belt, would be expected to reduce the incidence of EBV-related endemic Burkitt’s lymphoma, but not EBVnegative HIV-associated Burkitt’s lymphoma. However, most of the Burkitt’s lymphoma in the lymphoma belt remains EBV positive.5,8 It is important that further research into this disease should be done because it now occurs even more commonly in Africa than previously, and because elucidation of the pathogenesis of this rapidly growing tumour offers an unparalleled opportunity to unravel the mechanisms of carcinogenesis. Conflict of interest
I declare no conflicts of interest. References
1 de The G. Epstein-Barr virus and associated diseases. Course of medical virology, Institut Pasteur, 1995–96. Ann Intern Med 1997; 148: 357–66. 2 Haddow AJ. An improved map for the study of Burkitt’s lymphoma in Africa. East Afr Med J 1963, 40: 429–32. 3 van den Bosch CA, Hills M, Kazembe P, et al. Space-time caseclusters of endemic Burkitt’s lymphoma in Malawi. Leukemia 1993; 7: 1875–78. 4 Eltom MA, Jemal A, Mbulaiteye SM, et al. Trends in Kaposi’s sarcoma and non-Hodgkin’s lymphoma incidence in the United States from 1973 through 1998. J Natl Cancer Inst 2002; 94: 1204–10. 5 Otieno MW, Remick SC, Whalen C. Adult Burkitt’s lymphoma in patients with and without human immunodeficiency virus infection in Kenya. Int J Cancer 2001; 92: 687–91. 6 Joab I. Epstein-Barr virus and Burkitt’s lymphoma. J Trop Med 1999; 59: 499–502. 7 Magrath I, Jain V, Bhatia K. Epstein-Barr virus and Burkitt’s lymphoma. Semin Cancer Biol 1992; 3: 285–95. 8 Lazzi S, Ferrari F, Nyongo A, et al. HIV-associated malignant lymphoma in Kenya (equatorial Africa). Hum Pathol 1998; 29: 1285–89. 9 Carbone A. Emerging pathways in the development of AIDS-related lymphomas. Lancet Oncol 2003; 4: 22–29. 10 Iversen U, Iversen OH, Ziegler JL. Cell kinetics of African cases of Burkitt’s lymphoma: a preliminary report. Eur J Cancer 1972; 8: 305–08. 11 Burkitt D. The discovery of Burkitt’s lymphoma. Cancer 1983; 51: 1777–86. 12 Klein G. Lymphoma development in mice and humans: diversity of initiation is followed by convergent cytogenetic evolution. Proc Natl Acad Sci USA 1979; 76: 2442–46. 13 Bornkamm GW, Polack A, Eick D, et al. Chromosome translocations and Epstein-Barr virus in Burkitt's lymphoma. Onkologie 1987; 10: 196–204 (in German). 14 Haddow AJ. Age incidence in Burkitt’s lymphoma syndrome. East Afr Med J 1964; 41: 1–6. 15 Williams B. Tumor induction by viruses isolated from Mansonia africana. Afr J Med Sci 1981; 10: 19–28. 16 Precious SW. Isolation and persistence of Chikungunya virus in cultures of mouse brain cells. J Gen Virol 1974; 23: 271–79.
Oncology Vol 5 December 2004
17 Furstenberger G, Hecker E. On the active principles of the spurge family (Euphorbiaceae). 1. The skin irritant and tumor promoting diterpene esters of Euphorbia tirucalli L originating from South Africa. Z Naturforsch [C] 1985; 40: 631–46. 18 Haluska FG, Finver S, Tsujimoto Y, Croce CM. The t(8;14) chromosomal translocation occurring in B-cell malignancies results from mistakes in V-D-J joining. Nature 1986; 324: 158–61. 19 Thorley-Lawson DA, Gross A. Persistence of the Epstein-Barr virus and the origins of associated lymphomas. N Engl J Med 2004; 350: 1328–37. 20 Knowles DM, Cesarman E, Chadburn A. Correlative morphologic and molecular genetic analysis demonstrates three distinct categories of postransplantation lymphoproliferative disorders. Blood 1995; 85: 552–65. 21 Brichacek B, Davis J, Purtilo DT. In: Levine PH, Ablashi DV, Nonoyama M, et al (eds). Epstein Barr virus and human disease. Clifton: Humana Press, 1987: 53. 22 Henle G, Henle W, Klein G, et al. Antibodies to early Epstein-Barr virus induced antigens in Burkitt’s lymphoma. J Natl Cancer Inst 1971; 46: 861–71. 23 Pearson GR, Qualtire LF, Klein G, et al. Epstein-Barr virus-specific antibody-dependent cellular cytotoxicity in patients with Burkitt’s lymphoma. Int J Cancer 1979; 24: 402–06. 24 Rowe M, Rowe DT, Gregory CD, et al. Differences in B cell growth phenotype reflect novel patterns of Epstein-Barr latent gene expression in Burkitt’s lymphoma. EMBO J 1987; 6: 2743–51. 25 Yates J, Warren N, Reisman D, Sugden B. A cis-acting element from the Epstein-Barr virus genome that permits stable replication of recombinant plasmids in latently-infected cells. Proc Natl Acad Sci USA 1984; 81: 3806–10. 26 Snudden DK, Hearing J, Smith PR, et al. EBNA-1, the major nuclear antigen of Epstein-Barr virus, resembles “RCG” RNA binding proteins. EMBO J 1994; 13: 4840–47. 27 Srinivas SK, Sixbey JW. Epstein-Barr virus induction of recombinase-activating genes RAG1 and RAG2. J Virol 1995; 69: 8155–58. 28 Humme S, Reisbach G, Feederle R, et al. The EBV nuclear antigen, (EBNA1) enhances B cell immortalization several thousandfold. Proc Natl Acad Sci USA 2003; 100: 10989–94. 29 Drotar ME, Silva S, Barone E, et al. Epstein-Barr virus nuclear antigen-1 and Myc cooperate in development of lymphomas. Int J Cancer 2003; 106: 388–95. 30 Komano J, Maruo S, Kurozumi K, et al. Oncogenic role of EBVencoded RNAs in cell line. Akata J Virol 1999; 73: 9827–31. 31 Niller HH, Salamon D, Ilg K. The in vivo binding site for oncoprotein c-myc in the promoter for Epstein-Barr virus (EBV) encoding RNA (EBER) 1 suggests a specific role for EBV in development of lymphomas. Med Sci Monit 2003; 9: 1–9. 32 Greenwood BM, Oduloju AJ, Platts-Mill TAE. Partial characterization of a malaria mitogen. Trans R Soc Trop Med Hyg 1979; 73: 178. 33 Lam KMC, Syed N, Whittle H, Crawford DH. Circulating EpsteinBarr-carrying B cells in acute malaria. Lancet 1991; 337: 876–78. 34 Moss DJ, Burrows SR, Castelino DJ, et al. A comparison of EpsteinBarr virus T cell immunity in malaria endemic and non-endemic regions of Papua New Guinea. Int J Cancer 1983; 31: 727–32. 35 Bradley DJ, Warrell DA. Malaria epidemiology and immunity. In: Weatherall DJ, Ledingham JGG, Warrell DA (eds). Oxford textbook of medicine. 4th edn. Oxford: Oxford University Press, 1996: 838–42. 36 Nkhrumah FK, Perkins IV. Sickle cell trait, hemoglobin C trait and Burkitt’s lymphoma. Am J Trop Med Hyg 1976; 25: 633–36. 37 Coluzzi M, Calabro ML, Manno D, et al. Reduced seroprevalence of Kaposi’s sarcoma-associated Herpesvirus (KSHV), human herpesvirus 8 (HHV8), related to suppression of Anopheles density in Italy. Med Vet Entomol 2003; 17: 461–64. 38 Geser A, Brubaker G, Draper G. Effect of a malaria suppression program on the incidence of African Burkitt’s lymphoma. Am J Epidemiol 1987; 129: 740–52. 39 Haddow A. Epidemiological evidence suggesting an infective element in the aetiology. In: Burkitt DP, Wright DH (eds). Burkitt’s lymphoma. Edinburgh: Churchill Livingstone, 1970: 206–08. 40 Markoff L. Togaviridae: alphaviruses. In: Mandell GL, Bennett JE, Dolin R (eds). Principles and practice of infectious diseases. 4th edn. New York: Churchill Livingstone, 1995: 1455–59. 41 Williams EH, Smith PG, Day NE, et al. Space-time clustering of Burkitt’s lymphoma in the West Nile district of Uganda, 1961–75. Br J Cancer 1978; 37: 109–122.
http://oncology.thelancet.com
745
Personal view 42 Morrow RH, Pike MC, Smith PG, et al. Burkitt’s lymphoma: a timespace cluster of cases in Bwamba county of Uganda. BMJ 1971; 2: 491–92. 43 Pike MC, Williams EH, Wright B. Burkitt’s tumour in the West Nile district of Uganda, 1961–65. BMJ 1967; 2: 395–99. 44 van den Bosch C, Lloyd G. Chikungunya fever as a risk factor for endemic Burkitt’s lymphoma in Malawi. Trans R Soc Trop Med Hyg 1994; 94: 704–05. 45 Simpson DIH, Goma LKH, Williams EH. The West Nile survey. East Afr Virus Res Inst Rep 1964; 13: 36–39. 46 Williams MC, Simpson DIH, Woodhall JP. Work on Burkitt’s lymphoma syndrome-antibody studies using BUN. East Afr Virus Res Inst Rep 1964; 13: 40. 47 Dean A, Williams EH, Attobua G, et al. Clinical events suggesting herpes simplex infection before onset of Burkitt’s lymphoma: a case-control study in West Nile, Uganda. Lancet 1973; 2: 1225–28. 48 Ngu VA. Immunoglobulins in Burkitt’s lymphoma. Lancet 1966; 2: 414–16. 49 Yu MC, Tong MJ, Coursaget P, et al. Prevalence of hepatitis B and C, viral markers in black and white patients with hepatocellular carcinoma in the United States. J Natl Cancer Inst 1990; 82: 1038–41. 50 Sugawara Y, Makuuchi M, Kato N, et al. Enhancement of hepatitis C virus replication by Epstein-Barr virus-encoded nuclear antigen 1. EMBO J 1999; 18: 5755–60. 51 de Renzo A, Persicio E, de Marino F, et al. High prevalence of hepatitis G virus infection in Hodgkins disease and cell lymphoproliferative disorders: absence of correlation with Hepatitis C virus infection. Haematologica 2002; 87: 714–18. 52 Dutta U, Raina V, Garg PK, et al. A prospective study on the incidence of hepatitis B and C infections amongst patients with lymphoproliferatve disorders. Indian J Med Res 1998; 107: 78–82. 53 Chapin JP. Ecological aspects of bird distribution in tropical Africa. Am Naturalist 1923; 62: 106–25. 54 van den Bosch C, Griffin BE, Kazembe P, et al. Are plant factors a missing link in the evolution of endemic Burkitt’s lymphoma? Br J Cancer 1993; 68: 1232–35. 55 Osato T, Mizuno F, Imai S, et al. African Burkitt’s lymphoma and an Epstein-Barr virus-enhancing plant, E tirucalli. Lancet 1987; 1: 1257–58. 56 Mizuno F, Osato T, Imai S, et al. Epstein-Barr virus-enhancing plant promoters in East Africa. AIDS Res 1986; 2 (suppl 1): S151–55. 57 MacNeil A, Sumba OP, Lutzke ML, et al. Activation of the EpsteinBarr virus lytic cycle by the latex of the plant Euphorbia tirucalli. Br J Cancer 2003; 88: 1566–69. 58 Mizuno F, Koizumi S, Osato T, et al. Chinese and African Euphorbiaceae plant extracts: markedly enhancing effect on Epstein-Barr virus-induced transformation. Cancer Lett 1983; 19: 199–205. 59 Imai S, Sugiura M, Mizuno F, et al. African Burkitt’s lymphoma: a plant, Euphorbia tirucalli reduces Epstein-Barr virus-specific cellular immunity. Anticancer Res 1994; 14: 933–36.
746
Endemic Burkitt’s lymphoma
60 Steel CM, Morten JEN, Foster E. The cytogenetics of human B cell lymphoid malignancy: studies in Burkitt’s lymphoma and EpsteinBarr virus transformed lymphoblastoid cell lines. IARC Sci Publ 1985; 60: 265–92. 61 Aya T, Kinoshita T, Imai S, et al. Chromosomal translocations and c-myc activation by Epstein-Barr virus and Euphorbia tirucalli in B lymphocytes. Lancet 1991; 337: 1190. 62 Tanaka S, Southam CM. Joint action of West Nile virus and chemical carcinogens in production of papillomas in mice. J Natl Cancer Inst 1962; 29: 711–22. 63 Hall AJ, Wild CP. Liver cancer in low and middle income countries. BMJ 2003; 326: 994–95. 64 Ito Y, Yanase S, Ohigashi H, et al. Combined effect of the extracts from Croton tigium, Euphorbia lathyris or Euphorbia tirucalli and n-butyrate on Epstein-Barr Virus expression in human lymphoblastoid P3HR-1 and Raji cells. Cancer Lett 1981; 12: 175–80. 65 Hecker E. Cell membrane-associated protein kinase C as receptor of diterpene ester co-carcinogens of the tumour promoter type and the phenotypic expression of tumours. Arzneim Forsch Drug Res 1985; 35: 1890–903. 66 Goossens T, Klein U, Kuppers R. Frequent occurrences of deletions and duplications during somatic hypermutation: implications for oncogene translocations and heavy chain disease. Immunology 1998; 95: 2463–68. 67 Vanasse GJ, Concannon P, Willerford DM. Regulated genomic instability and neoplasia in the lymphoid lineage. Blood 1999; 94: 3997–4010. 68 Riboldi P, Ikematsu W, Brambilla B, et al. Diversity and somatic hypermutation of the Ig VHDJH, VkJK, and VJ gene segments in lymphoma B cells: relevance to the origin of the neoplastic B cell clone. Hum Immun 2003; 64: 69–81. 69 Buckley A, Gaidamovich S, Turchinskay A, Gould EA. Monoclonal antibodies identify the NS5 yellow fever virus non-structural protein in the nuclei of infected cells. J Gen Virol 1992; 73: 1125–30. 70 Seligman SJ, Gould EA. Live flavivirus vaccines : reasons for caution. Lancet 2004; 363: 2073–75. 71 Cheng WF, Hung CH, Chai CY, et al. Enhancement of sindbis virus self-replicating RNA vaccine potency by linkage of herpes simplex virus type 1 VP 22 protein to antigen. J Virol 2001; 75: 2368–76. 72 Anraku I, Harvey TJ, Linedale R, et al. Kunjin virus replicon vaccine vectors induce protective CD8+ T cell immunity. J Virol 2002; 76: 3791–99. 73 Hung C, He L, Juang J, et al. Improving DNA vaccine potency by linking Marek’s disease virus type 1 VP22 to an antigen. J Virol 2002; 76: 2676–82. 74 Ilyin KV, Kzhyshkowska JG, Imanova LR, et al. Type D retrovirus sequences in lymphocytes of the children with Burkitt-type lymphoma and their parents. Immunol Lett 2001; 78: 51–54. 75 Popescu NC. Genetic alterations in cancer as a result of breakage at fragile sites. Cancer Lett 2003; 192: 1–17.
Oncology Vol 5 December 2004
http://oncology.thelancet.com
