A Practical Approach to the New Lymphoma Classifications
Patrick A. Treseler, M.D., Ph.D.
Division of Surgical Pathology
Department of Pathology
University of California, San Francisco
NOTE: The following monograph is the property of the author and has been provided as a service to the medical community. No part of it may be reproduced without the written consent of the author.
Lymphoma classifications have evolved over the past 20 years from a scheme utilizing purely morphologic assessment (the Working Formulation); to a system that attempted to integrate morphologic, immunophenotypic, and genotypic data (the REAL Classification); and finally to a classification that hopes to integrate all these features into clinically relevant categories of equal use to pathologists and clinicians (the soon-to-be published WHO classification). Despite the proliferation of new names for lymphomas which these various revisions have provided, most of the newer classifications involve entities long recognized by pathologists, with the occasional splitting off (or recombining) of particularly entities based on newly recognized differences in morphology, phenotype, genotype, or behavior. Driving this continual reclassification of lymphoid neoplasms is not only new data concerning the biology of lymphoid neoplasia, but also continuing uncertainty as to what is most important in the accurate classification of lymphomas. Is it morphology, immunophenotype, genotype, clinical behavior, or something else? Obviously, the best system would be one which takes all such factors into account, but when discrepancies arise (as often happens in lymphoma cases), which factor should take precedence? Although modern pathology strives to understand diseases on the basis of genetics and molecular biology, genetic factors in lymphomagenesis are frequently poorly understood, and correlate poorly with clinical outcome in many cases.
A classic example is anaplastic large cell lymphoma (ALCL). Initially, ALCL was described as a neoplasm characterized by large pleomorphic lymphoid cells that expressed the lymphoid activation antigen CD30. When a subset of these tumors was found to contain a characteristic cytogenetic translocation, t(2;5), interest focused on this genetic defect as a possible defining element. It soon became apparent, however, that most such cases (which were generally of T-cell or null-cell phenotype) showed monomorphic rather than pleomorphic cytology (i.e., a "non-anaplastic" form of anaplastic large cell lymphoma) . Only later did it become clear that the t(2;5) abnormality could be found not only in tumors which were not anaplastic, but also in cases that were not CD30-positive, and occasionally in cases not even composed of large cells (the "small cell variant" of ALCL) . Furthermore, CD30-positivity was found to be present not infrequently in many different types of B-cell and T-cell lymphoma that shared little else in common with ALCL. Could a lymphoma be composed of cells which where CD30-positive, anaplastic, and large, and not be ALCL? What about the rare cases that had the t(2;5), but were clearly of B-cell phenotype and genotype? Were these also ALCL, like the more common T/null-cell cases? It all depends on which factor is considered most important morphology, immunophenotype, or genotype. A related issue is whether tumors with different genotypes and behavior always represent different disease entities, or whether these might better be regarded as prognostic features within a given entity in some cases. The new WHO classification will hardly end this continual reconsideration of the relative importance of various factors. It may however provide at least a temporary consensus concerning how specific entities are diagnosed.
B-cell Non-Hodgkin's Lymphomas
B-cell lymphomas are by far the most common non-Hodgkin's lymphomas in the United States. Despite the profusion of B-cell neoplasms recognized in the REAL and WHO classifications, most B-cell lymphomas can be correctly identified by morphologic examination alone if one can recognize a small number of basic lymphoid cell types. (B-cell neoplasms that tend to present with bone marrow and peripheral blood involvement (i.e., leukemias) will not be considered in this discussion, but are well-addressed in both the REAL and proposed WHO Classifications.)
Figure 1 above shows the eight basic lymphoid cell types found in B-cell lymphomas. Under each cell type is the name by which it has been known to many American pathologists, and, in parentheses under that, the name by which it would likely be known to most European pathologists using the Kiel classification. (With increasing usage of terminology of the Kiel classification in the REAL and WHO classifications, these latter terms are becoming more frequent in the U.S. literature.)
The basic initial distinction between the cell types is made on the basis of cell size, with the traditional yardstick being the average size of the nucleus of a phagocytic histiocyte. (Endothelial cell nuclei are about the same size, and are often easier to find.) Lymphoid cells the same size as histiocytic nuclei are considered intermediate in size. Those smaller are small cells, and those larger are large cells. (The confusing use of the term "small non-cleaved cell" for a lymphoid cell of intermediate size dates back to Lukes and Collins, who did not recognize an intermediate size group, and lumped intermediate size cells in the small cell category.)
A glance at Figure 1 will reveal that both small and intermediate size lymphoid cells have relatively scant cytoplasm; abundant cytoplasm is the mark of a large lymphoid cell. Small lymphoid cells all have typical mature lymphoid chromatin as their hallmark; all larger lymphoid cells have either coarsely clumped vesicular chromatin or, in the case of lymphoblasts, much finer nuclear chromatin. Thus it is not even really necessary to compare lymphoid cells to histiocytic or endothelial cell nuclei to judge their size. If the cytoplasm is abundant, the cell is large. If the cytoplasm is scant, mature lymphoid chromatin marks the small cells. Everything else is an intermediate size cell. (This system actually works better than true comparative measurement, since some lymphoblasts and small non-cleaved cells may actually be smaller than histiocytic or endothelial cell nuclei. Thus the adage "Never judge cell size by cell size".) Confusion between small cells and intermediate size cells is possible by morphology alone. (Even Rapaport failed to distinguish lymphoblasts from centrocytes in his initial classification.) However, lymphomas derived from intermediate size cells are very high grade, and almost always display high mitotic and apoptotic rates. Small cell lymphomas usually display relatively low mitotic and apoptotic rates. A final thing to note from Figure 1 is that little separates the morphologic appearance of small non-cleaved cells, centroblasts, and immunoblasts, except the amount of cytoplasm, which can make distinctions between these cells difficult. This is reflected in difficulty in distinguishing between certain lymphomas, as we shall see below.
B-cell lymphomas composed of large cells are large cell lymphomas, according to both the REAL and proposed WHO Classifications, regardless of whether they are composed of centroblasts or immunoblasts. The proposed WHO classification does permit, however, optional subclassification into centroblastic and immunoblastic types, and makes reference to the Kiel classification criteria, which require >90% immunoblasts for the diagnosis of immunoblastic lymphoma. (The subclassification of large cell malignancies will be further discussed in the section on immunophenotypic diagnosis of lymphomas ahead.) B-cell lymphomas composed of a uniform population of small non-cleaved cells are Burkitt's lymphomas. (An entity known as high-grade Burkitt-like B-cell lymphoma exists (as a provisional entity in the REAL, and as a variant of Burkitt's in the WHO) to handle cases that display more pleomorphism or contain more large cells than typical Burkitts lymphoma.) Most B-cell lymphomas with lymphoblastic morphology are lymphoblastic lymphomas, although a "blastic" or "blastoid" variant of mantle cell lymphoma has been described. Markers of early B-cell lineage (TdT, CD34, CD10) are helpful in confirming the diagnosis of lymphoblastic lymphoma. Mantle cell lymphoma, even with "blastic" cytology, will show a mature B-cell phenotype with expression of surface immunoglobulin and other characteristic markers of mantle cell lymphoma (see below).
Such a direct correlation of cell type to tumor type is not seen in the small B-cell lymphomas, several of which may contain a mixture of small lymphoid cell types. Here, however, architectural features and cytology combined will generally give strong clues to the lymphoma type, which can be confirmed by immunophenotyping. Only four types of small B-cell neoplasms typically present as lymphomas (i.e., outside the bone marrow): small lymphocytic lymphoma, follicular lymphoma, mantle cell lymphoma, and marginal zone lymphomas.
Small lymphocytic lymphoma (SLL) is the name given to chronic lymphocytic leukemia (CLL) when it presents outside the bone marrow. SLL displays a characteristic "pseudonodular pattern" at low power due to the presence of proliferation centers containing increased numbers of large lymphoid cells. This feature is highly sensitive and specific for SLL. Outside these proliferation centers, most cells are small round mature lymphocytes, and small cleaved cells (centrocytes) are uncommon.
Follicular lymphomas generally have at least some areas of follicular architecture that can be seen at low power. High power examination will reveal a prominent population of small cleaved cells (or centrocytes), with variable numbers of large non-cleaved cells (or centroblasts). (If a small B-cell lymphoma with numerous centrocytes appears entirely diffuse, one should suspect a mantle cell or marginal zone lymphoma.)
Mantle cell lymphoma (MCL) generally displays a diffuse or vaguely nodular pattern at low power. High power will reveal a population of mature-appearing small lymphoid cells that vary from round to cleaved. The relative lack of admixed large cells is a further clue to the diagnosis.
Like MCL, marginal zone lymphomas (MZLs) are also diffuse lymphomas that generally contain a prominent population of centrocytes. In contrast to the cellular monotony of MCL, however, MZLs are typically quite polymorphous, with frequent admixed plasmacytoid and monocytoid small lymphocytes, large lymphocytes, and histiocytes, producing a picture that may resemble a benign inflammatory infiltrate at high power. MZLs may occur in the lymph nodes ("monocytoid B-cell lymphoma"), extranodal sites ("low grade B-cell lymphomas of MALT type" or "MALToma"), or in the spleen ("splenic marginal zone lymphoma").
Immunophenotyping is quite useful in confirming the diagnosis of small cell B-cell lymphomas. While all are positive for pan-B cell markers (e.g., CD20), only SLL and MCL usually show co-expression of CD5. SLL can be distinguished by its expression of CD23; MCL is generally CD23-negative. MCL is also unique among small cell lymphomas in its expression of cyclin D1, a protein product of the t(11;14) translocation which can be detected immunohistochemically. About 60% of follicular lymphomas express CD10, which is uncommon in other small B-cell lymphomas. MZLs typically show no co-expression of either CD5 or CD10.
With an understanding of lymphoma cytology and architecture, the B-cell lymphomas in the 1994 REAL classification (Table 1) can thus be readily diagnosed in most cases. Immunophenotyping with a pan-B cell marker such as CD20 is recommended in cases of intermediate and large cell lymphoma to exclude non-B-cell tumors, and lymphoblastic lymphoma will require markers to confirm early B-cell phenotype and exclude other blastic neoplasms.
SLL and follicular lymphoma can generally be diagnosed by morphology, but in other cases of suspected small B-cell lymphoma, markers to confirm the diagnosis are suggested as discussed above.
In published reports which describe the proposed new WHO Classification, there are relatively few significant changes in the B-cell lymphomas from the 1994 REAL Classification. Table 2 shows the most recent proposed list of WHO B-cell neoplasms (excluding those that typically present in the bone marrow, and are generally considered leukemias). The only major change is the elimination of high-grade Burkitt-like B-cell lymphoma, however this entity still exists as a morphologic variant of Burkitt's lymphoma. Several significant issues are addressed in the new WHO Classification concerning morphologic and clinical variants and subtypes of particular lymphomas, which will be briefly discussed.
While it lists diffuse large B-cell lymphoma (DLBCL) as single distinct disease entity, the proposed WHO Classification recognizes six morphologic variants (which differ largely in their appearance or phenotype) and three subtypes (which differ in clinical behavior) of DLBCL (Table 3). The recognized six morphologic variants of DLBCL are centroblastic, immunoblastic, T-cell/histiocyte-rich, lymphomatoid granulomatosis type, anaplastic large B-cell, and plasmablastic. Centroblastic lymphoma is the prototypic diffuse large B-cell lymphoma, composed predominantly of large non-cleaved lymphocytes with smaller nucleoli which are frequently multiple and peripheral in the nucleus. Immunoblastic lymphoma, in contrast, is composed mainly of immunoblasts with large, prominent, centrally-located nucleoli. The WHO Committee has called for the development of criteria for distinguishing centroblastic from immunoblastic lymphoma (and cites the Kiel Classification criteria of >90% immunoblasts required for immunoblastic lymphoma). There is some evidence from the Kiel group to suggest that the immunoblastic variant may have a worse prognosis than centroblastic lymphoma. If large numbers of admixed reactive small T-lymphocytes or histiocytes are present, the lesion may be diagnosed as T-cell/histiocyte-rich large B-cell lymphoma. Several reports describe an aggressive clinical course in these patients. The large B-cells in these cases may resemble classical Reed-Sternberg cells or L&H Reed-Sternberg variants, raising the differential diagnosis of classical Hodgkins disease (CHD) or nodular lymphocyte predominance Hodgkins disease (NLPHD). Lack of CD30 and CD15 staining will help exclude classical Hodgkins disease, and lack of nodular architecture and a paucity of background small B-cells argues against NLPHD. Lymphomatoid granulomatosis is a clinically distinctive type of T-cell/histiocyte-rich large B-cell lymphoma that presents in the pulmonary tree. Previously thought to be related to nasal T/NK-cell lymphoma due to the prominent angiocentricity seen frequently in both tumors, recent evidence indicates a large B-cell origin. Diffuse large B-cell lymphomas may express CD30 and/or contain the t(2;5) translocation characteristic of ALCL, and these may be considered the anaplastic large B-cell variant. Most of these cases resemble typical DLBCLs, and do not show morphologic features of ALCL. Whether these cases are related to T-cell ALCL, or differ significantly from other DLBCLs, is controversial. Such tumors often have complex hyperdiploid karyotypes, leading to speculation that the t(2;5) translocation may be a later, secondary event in lymphomagenesis in these cases. Plasmablastic lymphoma is a rare aggressive Epstein-Barr virus (EBV)-positive AIDS-related B-cell malignancy first described by Delecluse et al. in 1997. In these features, and in its morphology, it resembles the high grade primary effusion lymphoma (PEL) also seen in AIDS patients. However, the tumor occurs principally in the oral cavity, and appears to lack the virus KSHV/HHV-8, which is universally present in PEL.
The three subtypes of diffuse large B-cell lymphoma recognized in the proposed WHO Classification are mediastinal (thymic) large B-cell lymphoma, primary effusion lymphoma, and intravascular large B-cell lymphoma. Mediastinal (thymic) large B-cell lymphoma differs from other forms of DLBCL not only in its location, but also in its phenotype and behavior. The tumor is often surface immunoglobulin negative, which is uncommon in DLBCL. CD30 is often positive, raising the differential diagnosis of CHD and ALCL. Clinically, mediastinal large B-cell lymphoma is unusual in that it tends to preferentially involve females in the fourth decade, and relapses are often extranodal, involving sites such as GI tract, liver, ovaries, and kidneys. While early reports suggested a poor prognosis, some recent data suggest that cure rates may be similar to other large cell lymphomas if treatment is appropriate. Primary effusion lymphoma, also known as Primary body-cavity based AIDS-related lymphoma appears to comprise a distinct clinicopathologic disease entity, which differs in many respects from more typical AIDS-related B-cell lymphomas. It generally occurs in late stage AIDS patients with a history of Kaposis sarcoma who present with effusions but no clinically evident mass lesion. Like Kaposis sarcoma, the disease occurs primarily in homosexual males. The disease tends to remain confined to the involved body cavity without further dissemination, but is nonetheless associated with an extremely poor prognosis, in part because the patients are generally debilitated and tolerate chemotherapy poorly. The median survival in reported cases averages two to three months. The tumor has large pleomorphic cells, occasionally multinucleated, with variably prominent, usually multiple nucleoli, and deeply basophilic cytoplasm by Wright-Giemsa stain. This appearance is in contrast to more typical AIDS-related lymphomas that generally appear as more typical large centroblastic, immunoblastic, or Burkitts/Burkitt-like B-cell lymphomas. Immunophenotyping reveals a characteristic "null cell" phenotype, with absence of most pan-B and T-cell markers, including surface and cytoplasmic immunoglobulin, but with uniform expression of CD30 and plasma cell markers such as CD38. B-cell markers that persist on mature plasma cells, such as CD79a, are also frequently expressed. This immunophenotype, which is most suggestive of plasma cell differentiation, contrasts with the mature B-cell phenotype expressed by other AIDS-related lymphomas. Immunoglobulin genes are clonally rearranged by molecular studies, confirming B-cell lineage for these neoplasms. Other significant differences from typical AIDS-related lymphomas are evident on the genetic level, where primary effusion lymphomas are found to be uniformly positive for Epstein-Barr virus (compared with 30-60% for other AIDS lymphomas) and to have consistent germline configuration of the c-myc proto-oncogene (which is rearranged in nearly all other AIDS lymphomas). First described by Walts and coworkers in 1990, primary effusion lymphomas became the focus of considerable attention in 1995, when Cesarman and colleagues found these tumors to contain a newly recognized human herpes virus termed Kaposis sarcoma-associated virus (KSHV) or human herpes virus-8 (HHV-8), which is also found associated with the cells of Kaposis sarcoma and multicentric Castlemans disease. Unlike Kaposis sarcoma, where the virus may only be detectable by PCR (and established KS cell lines may even be virus negative), body cavity-based lymphoma cells contain KSHV/HHV-8 in very high copy numbers. Since co-infection with EBV is seen in virtually every case, there has been speculation that the two viruses may cooperate in neoplastic transformation. Finally, intravascular large B-cell lymphoma, previously known as malignant angioendotheliomatosis, is a rare lymphoma characterized by an unusual tendency to proliferate and remain within vascular spaces. Vessels in virtually any organ may be involved by the disease, and thus the presenting symptoms are commonly non-specific (e.g., FUO, mental status changes, skin rash). This, combined with the absence of a mass lesion, makes diagnosis difficult clinically. The prognosis is extremely poor, although this may be due to long delays in reaching the diagnosis in many cases.
The proposed WHO Classification also takes a stand on the entity known as high grade Burkitt-like B-cell lymphoma, and its relationship to both Burkitt's lymphoma and DLBCL. High-grade Burkitt-like lymphoma had been recognized in the REAL classification as an entity with morphologic features intermediate between Burkitt's lymphoma and DLBCL. Specifically, most cases tended to resemble Burkitts lymphoma, but had either more cellular pleomorphism or more admixed large cells than classical Burkitt's. Lack of CD10 (typically expressed in Burkitt's lymphoma) was felt to favor the diagnosis, but cases could be CD10-positive. In a preliminary 1997 report on the proposed WHO Classification, it was stated that such tumors in non-immunosuppressed adults appeared more closely related to DLBCL than Burkitt's lymphoma. Subsequently, pathologists recommended making high-grade Burkitt-like B-cell lymphoma a subcategory of DLBCL. Pediatric oncologists expressed the concern that in children, cases classified as Burkitt-like behave identically to Burkitt's lymphoma, and would be undertreated if treated like DLBCL. The oncologists argued for the preservation of a category for "tumors that should be treated (like Burkitt's lymphoma), that is, very high-grade tumors." Thus, Burkitt-like lymphoma was retained as a morphologic variant of Burkitt's lymphoma. WHO Committee members expressed the opinion that the gold standard for the diagnosis of Burkitt's lymphoma should be the presence of the t(8;14) translocation (or its light chain variants t(2;8) or t(8;22)) . This translocation results in c-myc gene rearrangement and activation, which in turn causes all cells being perpetually in cycle. In the absence of cytogenetic studies or Southern blot analyses for c-myc rearrangement, a proliferative fraction close to 100% (as measured by Ki-67/MIB1 staining) was felt to be an acceptable surrogate marker. High-grade Burkitt-like lymphoma, as a variant of Burkitt's lymphoma, was further defined as a lymphoma that 1) morphologically resembles Burkitt's lymphoma, but has more pleomorphism or large cells than classical Burkitt's lymphoma; and 2) a proliferation fraction >99%.
The pathologists of the WHO Classification Committee also recommended guidelines for the separation of follicular lymphoma into three grades based on the number of large cells (centroblasts) present. Based on criteria initially developed by Mann and Berard, it was recommended that follicular lymphomas be graded as follows:
Grade 1 (follicular small cleaved): 1-5 centroblasts/high-power field (hpf)
Grade 2 (follicular mixed small and large cell): 6-15 centroblasts/hpf
Grade 3 (follicular large cell): >15 centroblasts/hpf.
Committee members felt that the most important distinction was to separate grade 3 tumors because of higher relapse rates in patients with grade 3 tumors. It was noted to be often difficult for pathologists to discriminate between grade 1 and 2 tumors, but this distinction was felt to be of little clinical significance. As in the REAL Classification, it was suggested that the degree of follicular architecture be reported as follows:
Predominantly follicular: >75% follicular architecture
Follicular and diffuse: 25-75% follicular
Predominantly diffuse: <25% follicular.
Grade 1 follicular lymphomas which appear entirely diffuse should have immunophenotyping performed to confirm follicular origin, and exclude other entities such as mantle cell lymphoma; the term "diffuse follicle center lymphoma, grade 1" can be used for these rare cases to avoid the non sequitur of a "diffuse follicular lymphoma". Areas of diffuse large cell lymphoma should be clearly reported as such, and not as "follicular lymphoma, grade 3, follicular and diffuse", since the presence of any component of DLBCL warrants more aggressive therapy.
The significance of large lymphoid cells in MZLs such as extranodal low grade B-cell lymphomas of MALT type was also addressed by the WHO Committee members. They cite a recent study of patients with low grade MALTomas treated primarily with antibiotics which found that if large cells were increased in number (5-10%, with clusters of <20 cells), patients had a slight but significantly worse prognosis (WHO latest). Cases with confluent clusters or sheets of large cells (defined as >20 cells) were found to behave similar to cases of DLBCL with no low grade component. It was recommended these latter cases be diagnosed as diffuse large cell lymphoma arising in a background of low grade MALToma, and not as high grade MALToma, since the latter term risks confusion with its low grade counterpart. The final WHO Classification will provide criteria for grading, likely along the lines described above.
Post-transplant lymphoproliferative disorders (PTLDs) are recognized by the WHO Committee as unique B-cell proliferations which span the spectrum from benign to malignant, and which require their own classification scheme separate from other B-cell neoplasms. A recent Society for Hematopathology Workshop greatly simplified the classification of these disorders. PTLDs are EBV-induced B-cell proliferations that may arise following either solid organ or bone marrow transplantation. The Workshop report describes three main categories of PTLDs:
"Early lesions" occur primarily in children within three months of transplantation, commonly involving sites such as tonsils or lymph nodes rather than extranodal sites. They do not show architectural effacement of involved tissues, nor destructive tissue infiltration. Most are polyclonal and regress either spontaneously or with minimal reduction in immunosuppression, but they can be fatal in some cases. Polymorphic PTLDs are destructive lesions that efface the architecture of involved tissues, which may be lymph nodes or extranodal sites. Morphologically, they resemble a "diffuse mixed small and large cell lymphoma" such as a marginal zone lymphoma. Most are monoclonal, but clonality does not predict behavior. A variable number will regress with reduction of immunosuppression. Those that persist or progress despite such therapy may require treatment for lymphoma. Monomorphic PTLDs are virtually all monoclonal, and resemble diffuse lymphomas. Indeed, most meet criteria for diffuse large B-cell lymphoma and can be diagnosed as such, but it should be noted that in the clinical setting of the post-transplant state, they are also monomorphic PTLDs. Many monomorphic B-cell cases will show aberrant co-expression of T-cell markers such as CD43 or CD45RO. Rare cases are of T-cell type, and may be considered T-cell lymphomas. Most monomorphic PTLDs are treated initially with reduction of immunosuppression, because occasional cases will respond to such therapy with durable remissions. Others may require standard lymphoma therapy.
Finally, for mantle cell lymphoma, the WHO Committee recognized that certain morphologic variants (such as the mantle zone pattern) have been reported to have a better prognosis in some studies, while cases with a blastic or blastoid morphology (but a mantle cell immunophenotype or genotype) have been reported to have a worse prognosis. The Committee suggests recording such patterns for the purpose of future research, but stratification by morphologic features was not felt to be required for clinical diagnostic purposes at this time since no effective therapy currently exists for any form of mantle cell lymphoma.
T-cell Non-Hodgkin's lymphomas
T-cell non-Hodgkin's lymphomas are much less common and much less well understood than their B-cell counterparts, and morphologic assessment is much less helpful in classification. The T-cell lymphomas recognized by the 1994 REAL Classification are listed in Table 4. While lymphoblastic lymphoma of T-cell type generally displays blastic morphology, the majority of the post-thymic (peripheral) T-cell lymphomas have morphology that can vary from mature appearing small lymphocytes to intermediate size blast-like cells to pleomorphic large cells to combinations of all three. Often, a continuum of atypical cells is seen spanning the size spectrum. Some studies have suggested that, in general, peripheral T-cell lymphomas with significant numbers of large cells may be more aggressive than those composed principally of small cells. However, often the grade is more dependent on tumor type than the cell morphology. For example, hepatosplenic gamma-delta T-cell lymphomas are almost always clinically aggressive, even if composed entirely of small cells. Another general feature of the peripheral T-cell lymphomas is that many are clinicopathologic entities, occurring in specific patient populations and involving particular body sites. Others display characteristic surface markers that aid in diagnosis, or have associations with etiologic viruses that can be detected by clinical tests. This sort of clinical information is frequently of much greater utility in reaching a diagnosis than the tumor cell morphology. For example, nasal T/NK-cell lymphomas typically occur in Asian patients, almost always involve extranodal sites (particularly the nose and palate), and are EBV-positive and clinically aggressive. Intestinal T-cell lymphoma typically involves the jejunum of adults who often have a history of gluten-sensitive enteropathy, and is also clinically aggressive. Overall, peripheral T-cell lymphomas are among the most aggressive of the non-Hodgkins lymphomas. Five year survival is under 30%, placing it in the worst prognostic group in a recent study conducted by the NHL classification project. Details concerning specific types of T-cell non-Hodgkin's lymphomas are provided in several excellent recent reviews.
The WHO Committee suggests in their recent proposed classification scheme that cytologic sub-classification of the peripheral T-cell lymphomas (e.g., as per the Kiel Classification) not be required for clinical diagnostic purposes at this time. They instead stress the importance of clinical correlation and immunophenotyping in the proper diagnosis of these entities, as discussed above. The T-cell lymphomas in the proposed WHO Classification (Table 5) are largely unchanged from the REAL (Table 4). Hepatosplenic gamma-delta T-cell lymphoma and subcutaneous panniculitis-like T-cell lymphomas have moved from provisional to fully accredited disease entities, and ALCL-Hodgkins like has been dropped given lack of evidence that it is a real entity. Angiocentric lymphoma has been changed to extranodal NK/T-cell lymphoma, nasal type, to reflect its lack of angiocentricity in some cases, and its apparent NK-cell origin. Specific conclusions and recommendations regarding anaplastic large cell lymphoma and NK/T-cell lymphomas will be discussed in the section on lymphoma immunophenotyping ahead.
In the years since the publication of the REAL Classification, Hodgkin's disease has moved from a tumor of uncertain histogenesis to a neoplasm now felt to be of B-cell origin in the vast majority of cases. Hodgkin's disease is apparently unique among B-cell neoplasms in being derived from germinal center B-cells with non-productive immunoglobulin gene rearrangements, apparently representing failed attempts at B-cell antibody production. Under normal circumstances, such cells undergo apoptosis in the germinal center, and the mechanism by which Reed-Sternberg cells escape cell death remains a mystery. The interested reader is referred to several excellent reviews of this fascinating subject.
Hodgkin's disease was recognized many years ago to consist of two very different entities. Nodular lymphocyte predominance Hodgkin's disease (NLPHD) is a mature B-cell neoplasm that routinely expresses CD45, B-cell antigens, and surface immunoglobulin, and which displays a characteristic nodular architecture and minimally pleomorphic, lobated, "popcorn-like" Reed-Sternberg variants that are generally CD30 and CD15 negative. NLPHD typically follows a very indolent clinical course. Non-lymphocyte predominance forms of Hodgkin's disease (also termed "classical Hodgkin's disease" or CHD) contain highly pleomorphic classical Reed-Sternberg cells, which are generally CD30 and CD15 positive, and are usually CD45 and T-cell/B-cell antigen negative. The 1994 REAL Classification subdivided classical Hodgkin's disease into nodular sclerosis, mixed cellularity, lymphocyte depletion, and lymphocyte-rich classical types (the last category serving largely to distinguish cases of classical Hodgkin's disease with large numbers of background lymphocytes from the very different NLPHD (Table 6).
What's New in the WHO: Hodgkin's Disease
In the proposed WHO Classification, Hodgkin's disease is largely unchanged from the REAL, except for the conversion of the "provisional" category of lymphocyte-rich CHD to a permanent one. Because of the accumulated evidence of B-cell origin, pathologists on the WHO Committee recommended a name change from Hodgkin's disease to Hodgkin's lymphoma. Because of resistance from some committee members concerned about possible confusion with non-Hodgkin's lymphomas, the recommendation was made to permit both terms as acceptable synonyms for the time being. The Committee also recommended that criteria for grading nodular sclerosis Hodgkin's disease (NSHD) be clearly stated in the final document, given that some studies have demonstrated a better prognosis for grade 1 (few RS cells) vs. grade 2 (many RS cells) NSHD. The committee recommended that such grading not be required for clinical purposes in routine diagnosis, but clear criteria may permit better study of the significance of grade in the future.
1. National Cancer Institute sponsored study of classifications of non-Hodgkin's lymphomas: summary and description of a working formulation for clinical usage. The Non-Hodgkin's Lymphoma Pathologic Classification Project. Cancer 1982 49: 2112-35 (1982).
9. Delabie J; Vandenberghe E; Kennes C et al.: Histiocyte-rich B-cell lymphoma. A distinct clinicopathologic entity possibly related to lymphocyte predominant Hodgkin's disease, paragranuloma subtype. Am J Surg Pathol 1992 16: 37-48 (1992).
10. Delecluse HJ; Anagnostopoulos I; Dallenbach F et al.: Plasmablastic lymphomas of the oral cavity: a new entity associated with the human immunodeficiency virus infection. Blood 1997 89: 1413-20 (1997).
11. Engelhard M; Brittinger G; Huhn D et al.: Subclassification of diffuse large B-cell lymphomas according to the Kiel classification: distinction of centroblastic and immunoblastic lymphomas is a significant prognostic risk factor. Blood 1997 89: 2291-7 (1997).
13. Guinee D, Jr.; Jaffe E; Kingma D et al.: Pulmonary lymphomatoid granulomatosis. Evidence for a proliferation of Epstein-Barr virus infected B-lymphocytes with a prominent T-cell component and vasculitis. Am J Surg Pathol 1994 18: 753-64 (1994).
16. Harris NL; Jaffe ES; Diebold J et al.: The World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues. Report of the Clinical Advisory Committee meeting, Airlie House, Virginia, November, 1997. Ann Oncol 1999 10: 1419-32 (1999).
17. Harris NL; Jaffe ES; Diebold J et al.: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues: report of the Clinical Advisory Committee meeting-Airlie House, Virginia, November 1997. J Clin Oncol 1999 17: 3835-49 (1999).
18. Harris NL; Jaffe ES; Diebold J et al.: The World Health Organization classification of hematological malignancies report of the Clinical Advisory Committee Meeting, Airlie House, Virginia, November 1997. Mod Pathol 2000 13: 193-207 (2000).
19. Harris NL; Jaffe ES; Diebold J et al.: The World Health Organization classification of neoplastic diseases of the haematopoietic and lymphoid tissues: Report of the Clinical Advisory Committee Meeting, Airlie House, Virginia, November 1997. Histopathology 2000 36: 69-86 (2000).
20. Harris NL; Jaffe ES; Stein H et al.: A revised European-American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group [see comments]. Blood 1994 84: 1361-92 (1994).
22. Jaffe ES; Harris NL; Chan JKC et al.: Proposed World Health Organization Classification of neoplastic diseases of hematopoietic and lymphoid tissues. Am J Surg Pathol 1997 21: 114-121 (1997).
24. Jaffe ES; Harris NL; Diebold J et al.: World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues. A progress report. Am J Clin Pathol 1999 111: S8-12 (1999).
30. Kuppers R: Identifying the precursors of Hodgkin and Reed-Sternberg cells in Hodgkin's disease: role of the germinal center in B-cell lymphomagenesis. J Acquir Immune Defic Syndr 1999 21 Suppl 1: S74-9 (1999).
33. Lazzarino M; Orlandi E; Paulli M et al.: Treatment outcome and prognostic factors for primary mediastinal (thymic) B-cell lymphoma: a multicenter study of 106 patients. J Clin Oncol 1997 15: 1646-53 (1997).
41. Rappaport H: Tumors of the Hematopoietic System (Armed Forces Institute of Pathology, Washington, D.C. 1966).
42. Rimokh R; Magaud JP; Berger F et al.: A translocation involving a specific breakpoint (q35) on chromosome 5 is characteristic of anaplastic large cell lymphoma ('Ki-1 lymphoma'). Br J Haematol 1989 71: 31-6 (1989).
43. Stein H; Mason DY; Gerdes J et al.: The expression of the Hodgkin's disease associated antigen Ki-1 in reactive and neoplastic lymphoid tissue: evidence that Reed-Sternberg cells and histiocytic malignancies are de-rived from activated lymphoid cells. Blood 1985 66: 848-58 (1985).