Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Denis Burkitt is credited with the pioneering work that led in 1958 to the first description of the clinical features of this unique tumor, the delineation of its geographic distribution, and the introduction of novel treatment protocols with chemotherapy. The eponym Burkitt's lymphoma (BL) appropriately recognizes his enormous contributions. The initial descriptions of BL were of rapidly growing tumors in the jaws of children residing in the malarial belt of equatorial Africa and New Guinea. This particular endemic form of BL is typically associated with Epstein-Barr virus (EBV). Elsewhere in the world, the association with EBV is much more variable. In vitro studies with BL cell lines have been instrumental in advancing the field of cancer genetics and tumor biology. The list of accomplishments includes the original descriptions of EBV itself, the first descriptions of the viral requirements for B-cell immortalization, and the mapping of the MYC locus to chromosome 8.
The nomenclature for this lymphoma entity has changed over the years. In the Rappaport classification, BL was called undifferentiated lymphoma, Burkitt type. Lukes and Collins classified BL as small non-cleaved follicular center cell lymphoma. The working formulation for clinical use, which separated lymphomas on the basis of their survival characteristics, classified BL as a clinically high-grade lymphoma of the small non-cleaved cell type. Both the Rappaport classification and the working formulation separated undifferentiated or small non-cleaved cell lymphomas into Burkitt and non-Burkitt types. The French-American-British classification of acute leukemia included a category of B-cell acute lymphoblastic leukemia (ALL), also known as L3-ALL. As appeared later on introduction of flow cytometry and cytogenetics, in almost all cases, L3-ALL reflects BL in leukemic phase. In the updated Kiel classification, BL was included as a distinct entity. In the Revised European-American Lymphoma (REAL) classification, BL was similarly considered a distinct lymphoma subtype, but the REAL classification proposal also included a provisional category of high-grade B-cell lymphoma, Burkitt-like, or Burkitt-like lymphoma. The last category recognized the existence of a small number of cases that challenged the distinction between BL and diffuse large B-cell lymphoma (DLBCL). These so-called gray zone cases were included in a provisional category in the World Health Organization (WHO) classification of 2008 (unclassifiable B-cell lymphoma, with features intermediate between DLBCL and BL). In the updated WHO classification of 2016 and as discussed in this chapter, BL is further defined by it molecular features, leaving less space for gray zone or borderline cases. In spite of all these considerations and its borders, there is no doubt that BL is a distinct entity with three distinct clinical and epidemiologic types.
BL is defined by the WHO classification as a highly aggressive lymphoid neoplasm, often presenting at extranodal sites or as an acute leukemia composed of monomorphic, medium-sized B cells with basophilic cytoplasm and a high mitotic rate. Translocations involving the MYC oncogene on chromosome 8 at band 8q24 are a constant feature. The frequency of EBV infection varies according to the epidemiologic subtype of BL (see later; Table 24-1 ).
Feature | BL | HGBL-DH | HGBL, NOS | DLBCL |
---|---|---|---|---|
Architecture | Diffuse | Diffuse, may have follicular component | Diffuse, may have follicular component | Diffuse +/– follicular component |
Starry-sky pattern | Usually present | May be present | May be present | Usually absent |
Mitoses | Many | Variable | Variable | Variable |
Cytology | Monomorphic | DLBCL, intermediate BL-DLBCL, blastoid | Intermediate BL-DLBCL, blastoid; not DLBCL | Variable |
Nuclear shape | Round or slightly irregular | Like DLBCL, NOS; more pleomorphic, irregular, ovoid | Like BL; more pleomorphic, irregular, ovoid | Centroblasts predominate |
Nuclear size | Intermediate | Large or intermediate | Intermediate or variable | Large |
Nucleoli | Multiple (2-5), medium sized, basophilic | Variable; in blastoid, inconspicuous | Variable; in blastoid, inconspicuous | Centroblasts, 2-3 adjacent to nuclear membrane; immunoblasts, 1-2 central |
Cytoplasm | Deeply basophilic; often vacuolated; squared-off appearance | Variable; vacuoles often absent; may lack squared-off appearance | Variable; vacuoles often absent; may lack squared-off appearance | Variable; typically no or few vacuoles |
CD10 expression | Positive | BCL2 DH positive BCL6 DH often negative |
Variable | Variable (40%-60%) |
BCL6 expression | Positive | Usually positive | Usually positive | 70%-80% positive |
BCL2 expression | Negative or weak | BCL2 DH cases strongly positive BCL6 DH cases often negative or weak |
Variable | 60%-70% positive, mostly in ABC subtype; in GCB subtype, if BCL2 rearrangement positive |
MYC expression | Mostly >80% | Mostly >30% | Unknown | Variable; if MYC rearrangement, >30% |
Ki67 proliferation index | >95% | Variable, may be <50% | Variable | Variable |
EBV-EBER | Positive (15%-30%) * | Negative or not studied | Negative or not studied | Usually negative |
Cytogenetics | t(8;14), t(2;8), or t(8;22) | By definition, 8q24 plus 18q21 or 3q27 breakpoint, t(8;14) or variant in 60%; in 35%-40%, non–IG/ MYC | By definition, no DH; may have isolated 8q24, 3q27, or 18q21 breakpoints also with gain or amplification of other loci | By definition, no DH; 3q27 breakpoint in 30%, 18q21 breakpoint in 15%-20%, 8q24 breakpoint in 5%-10% |
Complexity † | Low; in >50%, only t(8;14) or variant | Often complex | Often complex | Often complex |
Other molecular features | Activating ID3 >> TCF3 (biallalic) mutations in >50% | TP53 mutations frequent in MYC-BCL2 DH and blastoid cases; monoallelic ID3 mutations may occur | Not studied | Different mutations in the various molecular subsets (GCB versus ABC type); see Chapter 23 |
* Percentages apply to sporadic BL.
† Complex karyotype is defined by three or more clonal aberrations (including the 8q24 breakpoint) as detected by karyotyping.
Three clinical variants of BL are recognized, with substantial differences in clinical presentation and anatomic localization of the primary tumor, subtle differences in morphology, and variable molecular genetics and biology.
BL is endemic in the malaria belt of equatorial Africa, which stretches from Senegal and Mauritania in the northwest to Tanzania and Mozambique in the southeast, and in Papua New Guinea. The areas of sub-Saharan Africa most implicated are those at altitudes with high annual rainfall and elevated temperatures, corresponding to regions endemic for malaria. People living in urban areas are largely spared from BL. Endemic BL affects primarily young children, with a peak incidence in those aged 4 to 10 years and a 2 : 1 male-to-female predominance. At this age, the tumors are most often extranodal, particularly involving the jaw, facial bones, and orbit. The majority of cases of endemic BL are EBV + and demonstrate distinct molecular rearrangements of both MYC and one of the three immunoglobulin (IG) genes ( Table 24-2 ).
Feature | Endemic BL | Sporadic BL | AIDS-Associated BL |
---|---|---|---|
Predominant MYC breakpoint in t(8;14)(q24;q32) | Far 5′ (centromeric) of MYC (class III) | Exon and intron 1 (class I) and 5′ (centromeric) of MYC (class II) | Exon and intron 1 (class I) |
Predominant IGH breakpoint in t(8;14)(q24;q32) | VDJ region | Switch region | Switch region |
Somatic IGH mutations | Yes | Yes | Yes |
EBV positivity | >90% | 5%-30% | 25%-40% |
This variant is seen throughout the world, commonly afflicting children and young adults. It accounts for 30% to 50% of pediatric lymphomas but is uncommon in adults, thus accounting for only 1% to 2% of all lymphomas in Western Europe and North America. Sporadic BL is uncommon in those younger than 2 years and has a peak between 5 and 15 years. The median age of adult patients is approximately 30 years, but there is some discussion about the incidence after this age; some studies report a steady increase up to 65 years or a trimodal age-specific pattern. Thus, incidental cases may present in patients older than 65 years. There is a male predominance of about 2 : 1 to 5 : 1, which seems to be more prominent in children and in particular in children with involvement of the face and jaws. The majority of the sporadic cases presents with abdominal masses. The ileocecal region is the most frequent site of involvement, whereas involvement of the jaw or the orbit is unusual. Similarly to endemic BL, ovaries, kidneys, and breasts may also be involved. Breast involvement (often bilateral and massive) has been associated with onset during puberty, pregnancy, or lactation. EBV is seen in less than 30% of the reported cases and in only 10% to 20% in most Western countries. Sporadic BL can occur in those living in endemic regions and may account for some cases with atypical presentations or lack of EBV involvement. At the molecular level, the breakpoint sites for both IGH and MYC are usually different from those encountered in endemic BL (see the section on genetics, Table 24-2 , and Fig. 24-9 ).
This variant of BL is seen primarily in association with human immunodeficiency virus (HIV) infection and accounts for approximately 30% to 40% of all acquired immunodeficiency syndrome (AIDS)–related cases of lymphoma, but the relative incidence of BL in patients with AIDS has been falling steadily in recent years, perhaps influenced by increasing use of highly active antiretroviral therapy (HAART). Infection of BL cells with EBV is seen in 25% to 40% of these cases. Many cases demonstrate slightly atypical cytologic features and often exhibit plasmacytoid differentiation. This variant of BL may also occur in other immunodeficiency states, including congenital disorders such as Duncan’s disease (X-linked lymphoproliferative disease associated with SH2D1A mutations) and iatrogenic causes, such as the immunosuppression required after organ transplantation. However, BL is uncommon in the latter setting.
In endemic BL, massive early childhood EBV infection and the chronic antigenic stimulation and immune suppression that accompany persistent malarial infection have been suggested as causative factors. HIV infection has been similarly implicated in immunodeficiency-associated BL. Other cofactors, such as exposure to Euphorbia tirucalli in endemic BL and pesticide exposure in both adults and children, have been noted.
The most consistent factor implicated in the pathogenesis of BL is translocation of the MYC oncogene. The origin of the translocations and mutations in MYC are discussed later under Genetics. Older data from the literature suggested that MYC has two major roles in B cells: promoting cellular proliferation and downregulating the expression of human leukocyte antigen (HLA) class I molecules, thus allowing the tumor cells to evade host immune control. Subsequent data showed that MYC overexpression not only enhances cellular proliferation but also promotes apoptosis. More recent work shows that the role of MYC is much more complex (see the section on genetics). Most important, except for a minor population of approximately 10% of the cells, MYC is not expressed in normal germinal center B cells. This is in great contrast with BL, in which almost all lymphoma cells show a distinct overexpression as assessed by immunohistochemistry.
Although a consistent finding in BL, MYC overexpression is insufficient on its own to induce lymphoma. Overexpression of MYC in transgenic mice leads to a polyclonal expansion of precursor B cells, but monoclonal neoplasms do not develop for 6 to 9 months and then almost certainly as the result of additional genetic insults. In aggregate, these data suggest that although MYC is important for the development of BL, it may be only an initiating event that requires additional genetic alterations to express the full malignant phenotype. Indeed, a number of genetic alterations (to be discussed later) may conspire to subvert the apoptotic signaling associated with MYC , including mutations in MYC itself, in the p53-MDM2-ARF pathway, and in TCF3 and its inhibitor ID3 .
EBV infection represents another important factor in the pathogenesis of BL. EBV is a nearly ubiquitous human herpesvirus that is capable of either transforming B cells or persisting within these cells in a latent state. Latent EBV infection is characterized by three different patterns of gene expression. In latency pattern type I, which is characteristic of BL, the viral-associated genes expressed are EBER-1 , EBER-2 , and EBNA-1 . In latency pattern II, which is characteristic of Hodgkin's lymphoma, peripheral T-cell lymphomas, and primary effusion lymphomas in the HIV setting, the viral genes expressed include EBER-1 , EBER-2 , EBNA-1 , LMP-1 , LMP-2A , and LMP-2B . In the type III latency pattern, seen in post-transplantation lymphoproliferative disorders and lymphoblastoid cell lines, the full spectrum of latent viral genes is expressed: EBER-1 and EBER-2 ; EBNA-1 , 2 , 3A , 3B , and 3C ; and LMP-1 , 2A , and 2B .
Latency type I pattern is present in the majority of endemic BL. However, in these regions, EBV is not present in all cases, raising the possibility that other factors like HIV infection may play a role. Moreover, there is a gradually decreasing level of EBV infection in the countries around equatorial Africa to a level of 20% to 30% in sporadic BL cases, which is also difficult to reconcile with EBV's being a prerequisite for BL.
In support of EBV's role as an important causative agent in BL, most studies have shown evidence of a clonal infective event. With use of probes to the terminal repeat sequences of EBV, BL cells harbor latent virus in the form of episomal DNA in keeping with viral infection before clonal expansion. Contrasting with these data are in vitro studies of sequential clinical samples and corresponding cell lines in which EBV infection was shown to be a later event. The two main theories concerning the role of EBV in the pathogenesis of BL have much to do with the timing of infection in relation to translocation of the MYC oncogene. In one hypothesis, EBV infection induces a polyclonal expansion of B cells, and the increased mitotic activity increases the likelihood of inducing an aberrant MYC rearrangement. Under these conditions, EBV may only potentiate tumor development by stimulating cellular proliferation, decreasing apoptotic signaling, or increasing genetic instability. The “hit and run” model, in which the virus does its damage and then is lost (or most of its oncogenic potential is lost), fits with this concept and could also explain why some cases of endemic BL are negative for latent EBV genomes.
The second hypothesis suggests that infection by the virus occurs after the MYC translocation. This theory can be reconciled more easily with the latency type I EBV gene expression seen in BL, consisting of only EBER-1 , EBER-2 , and EBNA-1 (latency type I). This pattern of gene expression may offer a survival advantage, allowing an escape from immune surveillance because EBNA-1 does not induce a potent cytotoxic T-cell response. Other data suggest that MYC and EBNA-1 may cooperate in promoting lymphomagenesis.
An important observation is that chronic malaria infection and higher blood EBV loads are associated with shifts in B-cell populations as assessed by flow cytometry and by a higher activation-induced cytidine deaminase (AID) expression, thereby facilitating chromosomal breaks in children living in these endemic regions. These observations and the relationship with the genesis of BL are also supported by experimental work in mice infected with malaria. Overexpression of E2F1 may also play a part in cell cycle deregulation in BL. In spite of these studies, the precise role of holoendemic malaria and HIV infection as well as of EBV infection as pathogenetic factors has yet to be established.
The findings associated with the extremely rapid growth and massive apoptosis of neoplastic cells typify virtually all cases of BL. Serious oncologic emergencies may develop, including bowel intussusception and perforation, ureteric obstruction, and paraplegia secondary to paraspinal masses with cord compression but also metabolic events due to tumor lysis. Involvement of the central nervous system (CNS) seems to be a poor prognostic sign in all three forms of BL. Some unique clinical findings are associated with the different epidemiologic subtypes.
Endemic BL involves extranodal sites, particularly the jaw and orbit, in approximately 50% of cases, in particular in boys. These structures are involved in approximately 70% of cases in children younger than 5 years but in only 25% of those older than 14 years ( Fig. 24-1 ). Other sites of involvement include the distal ileum, cecum, ovary, kidney, and breast.
Facial involvement by lymphoma may fill the sinuses or cause loosening of the teeth. Gross abnormalities of the orbit are common, as is involvement of cranial nerves. Involvement of the bone marrow or presentation as acute leukemia is uncommon in endemic BL. Patients may complain of abdominal pain, swelling, or a change in bowel habits.
In sporadic BL, involvement of the facial structures, in particular the jaw, is uncommon. Eighty percent to 90% of cases present with involvement of intra-abdominal structures. The ileocecal region is most commonly involved; but in some cases, the tumor may be restricted to the appendix, ascending colon, or peritoneum. Other sites include the ovary, kidney, and breast. Of note, bilateral breast involvement is associated with the onset of puberty, pregnancy, or lactation. Patients may have pleural effusions. Bone marrow involvement is more common than in endemic BL and is seen at some point in the course of the disease in almost every fatal case. Although CNS involvement is uncommon at diagnosis, it develops eventually in most cases of sporadic BL unless effective CNS chemotherapy is given along with curative systemic chemotherapy. Peripheral lymph node involvement is seen in only 10% to 15% of cases and is more common in adults than in children. Involvement of Waldeyer's ring and in particular the mediastinum is rare. Patients with advanced-stage or bulky disease may have circulating neoplastic cells identified in the peripheral blood smear. A diagnosis of BL is reserved for those cases with greater than 25% bone marrow involvement at diagnosis. Importantly, very infrequently patients are seen with a leukemic clinical picture with massive bone marrow involvement of otherwise immunophenotypically and genetically classic BL, but without any apparent other extranodal involvement.
BL may be related to primary and acquired immunodeficiencies, AIDS being by far the most important. According to the older literature, approximately one third of the lymphomas seen in association with AIDS are BLs. The appearance of BL in an individual with HIV infection is an AIDS-defining illness. Factors associated with the development of immunodeficiency-related BL are younger age (10 to 19 years where it shows a bimodal age distribution not seen in the other lymphomas associated with HIV infection, in particular EBV + immunoblastic lymphoma); relatively high or intermediate CD4 cell counts (in fact, the risk for BL peaks at CD4 lymphocyte counts well above the laboratory cutoff for AIDS onset of 200 cells/µL); and long-standing HIV infection, especially in patients not receiving HAART. AIDS-related BL is often advanced at diagnosis, with B symptoms and poor performance status present in at least 70% of cases. Although nodal involvement is more common than in endemic BL, extranodal involvement remains frequent, usually involving the gastrointestinal tract, bone marrow, or CNS. Prognosis is linked to both the extent of the tumor volume and the severity of the underlying immunodeficiency. Important adverse prognostic factors include low CD4 count, presence of AIDS before the development of BL, poor performance status, and, with respect to the lymphoma stage IV disease, especially CNS or bone marrow involvement. The relationship with the underlying immune compromise is supported by the fact that the incidence of all AIDS-related lymphomas, including BL, has dropped dramatically since HAART became widely available. Recent reports have emphasized the need to distinguish between AIDS-related BL and AIDS-related DLBCL for treatment purposes because CHOP (cyclophosphamide, hydroxydaunomycin, Oncovin, prednisone)–based regimens are less effective than in AIDS-related DLBCL. However, infusional chemotherapy regimens, such as EPOCH (etoposide, prednisone, Oncovin, cyclophosphamide, hydroxydaunomycin) with rituximab, have shown a high efficacy for all HIV-associated aggressive B-cell lymphomas.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here