The World Health Organization Classification of Hematological

Home , Granulomatous slack skin, List of diseases (A), List of diseases (L), List of diseases (O), List of diseases (T), List of diseases (W)

SPECIAL ARTICLE

The World Health Organization Classification of Hematological Malignancies Report of the Clinical Advisory Committee Meeting, Airlie House, Virginia, November 1997 Nancy Lee Harris, M.D., Elaine S. Jaffe, M.D., Jacques Diebold, M.D., Georges Flandrin, M.D., H. Konrad Muller-Hermelink, M.D., James Vardiman, M.D., T. Andrew Lister, M.D., Clara D. Bloomfield, M.D. Department of Pathology, Massachusetts General Hospital and Harvard Medical School (NLH), Boston, Massachusetts; National Cancer Institute (ESJ), Bethesda, Maryland; Hotel Dieu (JD) and Hopital Necker (GF), Paris, France; University of Wurzburg (HKM-H), Wurzburg, Germany; Pritzker School of Medicine, University of Chicago (JV), Chicago, Illinois; the Department of Medical Oncology, St. Bartholomew’s Hospital (TAL), London, England; and the Ohio State University Comprehensive Cancer Center (CDB), Columbus, Ohio

phoid Neoplasms to myeloid and histiocytic neo- Since 1995, the European Association of Patholo- plasms. The classification of myeloid neoplasms gists and the Society for Hematopathology have recognizes distinct entities defined by a combina- been developing a new World Health Organization tion of morphology and cytogenetic abnormalities. (WHO) classification of hematologic malignancies. The Clinical Advisory Committee meeting, which The classification includes lymphoid, myeloid, his- was organized around a series of clinical questions, tiocytic, and mast cell neoplasms. was able to reach a consensus on most of the ques- The WHO project involves 10 committees of pa- tions posed. The questions and the consensus are thologists, who have developed lists and definitions discussed in detail in this article. Among other of disease entities. A Clinical Advisory Committee of things, the Clinical Advisory Committee concluded international hematologists and oncologists was that clinical grouping of lymphoid neoplasms was formed to ensure that the classification will be use- neither necessary nor desirable. Patient treatment ful to clinicians. A meeting was held in November is determined by the specific type of lymphoma, 1997 to discuss clinical issues related to the classi- with the addition of grade within the tumor type, if fication. The WHO has adopted the Revised applicable, and clinical prognostic factors such as European-American Classification of Lymphoid the international prognostic index. Neoplasms, published in 1994 by the International The experience of developing the WHO classifica- Lymphoma Study Group, as the classification of lymphoid neoplasms. This approach to classification has produced a new and exciting degree of tion is based on the principle that a classification is cooperation and communication between oncolo- a list of “real” disease entities, which are defined by gists and pathologists from around the world. This a combination of morphology, immunophenotype, should facilitate progress in the understanding and genetic features, and clinical features. The relative treatment of hematologic malignancies. importance of each of these features varies among diseases, and there is no one “gold standard.” The KEY WORDS: Classification, Histiocytic, Leukemia, WHO classification has applied the principles of the Lymphoma, Mast cell, Myeloid. Revised European-American Classification of Lym- Mod Pathol 2000;13(2):193–207

The Society for Hematopathology and the Euro- Copyright © 2000 by The United States and Canadian Academy of pean Association of Hematopathologists have un- Pathology, Inc. dertaken as a joint project the development of a VOL. 13, NO. 2, P. 193, 2000 Printed in the U.S.A. Date of acceptance: November 2, 1999. classification of hematologic neoplasms for the This article also is being published simultaneously in the Journal of World Health Organization (WHO). A steering com- Clinical Oncology, Annals of Oncology, The Hematology Journal, and His- topathology. mittee composed of members of both societies has Address reprint requests to: Nancy Lee Harris, M.D., Pathology, Warren 2, been formed, and 10 committees have been as- Massachusetts General Hospital, Fruit Street, Boston, MA 02114; e-mail: [email protected]; fax: 617-726-7474. signed the task of arriving at a consensus list of

193 myeloid, lymphoid, and histiocytic neoplasms, with TABLE 1. Proposed WHO Classification of Myeloid Neoplasms descriptions and criteria for diagnosis. A new classification for lymphoid neoplasms was recently Myeloproliferative Diseases (MPD) Chronic myelogenous leukemia, Philadelphia chromosome (Ph1) proposed (1), and the goals of the WHO project are [t(9;22)(qq34;q11), BCR/ABL]ϩ to update and revise that classification, with input Chronic neutrophilic leukemia from additional experts to broaden the consensus, Chronic eosinophilic leukemia/hypereosinophilic syndrome Chronic idiopathic myelofibrosis and to extend the principles of disease definition Polycythemia vera and consensus building to the myeloid and histio- Essential thrombocythemia cytic neoplasms. More than 50 pathologists from Myeloproliferative disease, unclassifiable Myelodysplastic/Myeloproliferative Diseases around the world have been involved in the project Chronic myelomonocytic leukemia (CMML) since 1995. Proponents of all major lymphoma and Atypical chronic myelogenous leukemia (aCML) leukemia classifications have agreed that if a rea- Juvenile myelomonocytic leukemia (JMML) Myelodysplastic Syndromes (MDS) sonable consensus emerges from this effort, they Refractory anemia (RA) will accept the WHO as the standard classification with ringed sideroblasts (RARS) of hematologic malignancies. without ringed sideroblasts Refractory cytopenia (myelodysplastic syndrome) with multilineage The proposed WHO classification of hematologic dysplasia (RCMD) malignancies stratifies these neoplasms primarily Refractory anemia (myelodysplastic syndrome) with excess blasts according to lineage: myeloid neoplasms, lymphoid (RAEB) 5q- syndrome neoplasms, mast cell disorders, and histiocytic neo- Myelodysplastic syndrome, unclassifiable plasms (Tables 1– 5). Within each category, distinct Acute Myeloid Leukemias (AML)a diseases are defined according to a combination of Acute myeloid leukemias with recurrent cytogenetic translocations AML with t(8;21)(q22;q22), AML1(CBF␣)/ETO morphology, immunophenotype, genetic features, Acute promyelocytic leukemia (AML with t(15;17)(q22;q11-12) and and clinical syndromes. The relative importance of variants, PML/RAR␣) each of these criteria differs among the neoplasms, AML with abnormal bone marrow eosinophils (inv(16)(p13q22) or t(16;16)(p13;q11), CBF␤/MYH11X) and there is no one gold standard for classification AML with 11q23 (MLL) abnormalities of all hematologic malignancies. The goal is to de- Acute myeloid leukemia with multilineage dysplasia fine disease entities that can be recognized by pa- with prior myelodysplastic syndrome without prior myelodysplastic syndrome thologists and that have clinical relevance. Acute myeloid leukemia and myelodysplastic syndrome, therapy To ensure that the proposed classification will be related of maximum use to oncologists, the Steering Com- Alkylating agent related Epipodophyllotoxin related (some may be lymphoid) mittee invited expert hematologists and oncologists Other types to form a Clinical Advisory Committee (CAC), with Acute myeloid leukemia (AML) not otherwise categorized American and European co-chairs. The charge to AML minimally differentiated AML without maturation the committee was to review the proposed classifi- AML with maturation cation and advise the pathologists on its clinical Acute myelomonocytic leukemia utility. More than 40 hematologists and oncologists Acute monocytic leukemia Acute erythroid leukemia from around the world agreed to participate. The Acute megakaryocytic leukemia proposed classification was circulated, and all par- Acute basophilic leukemia ticipants were invited to submit topics and ques- Acute panmyelosis with myelofibrosis Acute Biphenotypic Leukemias tions for discussion. A meeting was held in Novem- Only major disease categories are listed; subtypes and variants will be ber 1997 at Airlie House, Virginia, to which the CAC discussed in detail in the text. and all pathologists involved in the WHO commit- a Acute lymphoid leukemias are included under lymphoid neoplasms tees, as well as the Executive Committees of the two and in Table 7. hematopathology societies, were invited. The meeting was organized around a series of questions, developed from those submitted by CAC show of hands was taken as a vote. Following the members as well as those posed by the pathologists. meeting, a poll of the participants, as well as several Only issues that were controversial were discussed; additional meetings of the pathology Steering Com- diseases for which there were no new questions or mittee and the co-chairs of the CAC, was held data were accepted as previously defined. Only to resolve residual questions. The final classifica- lymphoid and myeloid neoplasms were discussed tion will be published under the auspices of the at this meeting; histiocytic and mast cell tumors WHO (2). were not considered. Participants were invited to present data relevant to each question, and open MYELOID NEOPLASMS discussion followed. At the end of each session, the clinicians present were asked to arrive at a consen- Although there have been many advances in the sus regarding each question (as well as on other understanding of genetic factors in the biology of issues raised at the meeting); when necessary, a the myeloid neoplasms, particularly the acute leu-

194 Modern Pathology TABLE 2. Proposed WHO Classification of Lymphoid TABLE 3. Categories of Posttransplant Neoplasms Lymphoproliferative Disorders (PTLD)

B-Cell Neoplasms Early lesions Precursor B-cell neoplasm Reactive plasmacytic hyperplasia Precursor B-lymphoblastic leukemia/lymphoma (precursor B-cell Infectious mononucleosis-like acute lymphoblastic leukemia) PTLD polymorphic Mature (peripheral) B-cell neoplasmsa Polyclonal (rare) B-cell chronic lymphocytic leukemia/small lymphocytic lymphoma Monoclonal B-cell prolymphocytic leukemia PTLD monomorphic (classify according to lymphoma classification) Lymphoplasmacytic lymphoma B-cell lymphomas Splenic marginal zone B-cell lymphoma (Ϯ villous lymphocytes) Diffuse large B-cell lymphoma (Immunoblastic, Centroblastic, Hairy cell leukemia Anaplastic) Plasma cell myeloma/plasmacytoma Burkitt/Burkitt-like lymphoma Extranodal marginal zone B-cell lymphoma of MALT type Plasma cell myeloma Nodal marginal zone B-cell lymphoma (Ϯ monocytoid B cells) T-cell lymphomas Follicular lymphoma Peripheral T-cell lymphoma, not otherwise categorized Mantle cell lymphoma Other types (Hepatosplenic, gamma-delta, T/NK) Diffuse large B-cell lymphoma Other types (rare) Mediastinal large B-cell lymphoma Hodgkin disease–like lesions (associated with methotrexate therapy) Primary effusion lymphoma Plasmacytoma-like lesions Burkitt lymphoma/Burkitt cell leukemia T and NK-Cell Neoplasms Precursor T-cell neoplasm Precursor T-lymphoblastic lymphoma/leukemia (precursor T-cell TABLE 4. Mast Cell Diseases acute lymphoblastic leukemia) Cutaneous mastocytosis Mature (peripheral) T-cell neoplasms** Systemic mast cell disease (Ϯ skin involvement) T-cell prolymphocytic leukemia Systemic mast cell disease with associated hematologic disorder (Ϯ skin T-cell granular lymphocytic leukemia involvement) Aggressive NK-cell leukemia Mast cell leukemia/sarcoma Adult T-cell lymphoma/leukemia (HTLV1ϩ) Extranodal NK/T-cell lymphoma, nasal type Enteropathy-type T-cell lymphoma TABLE 5. Histiocytic and Dendritic Cell Neoplasms Hepatosplenic ␥␦ T-cell lymphoma Subcutaneous panniculitis-like T-cell lymphoma Macrophage/Histiocytic neoplasm Mycosis fungoides/Sezary syndrome Histiocytic sarcoma Anaplastic large cell lymphoma, T/null cell, primary cutaneous type Dendritic-cell neoplasms Peripheral T-cell lymphoma, not otherwise characterized Langerhans cell histiocytosis Angioimmunoblastic T-cell lymphoma Langerhans cell sarcoma Anaplastic large cell lymphoma, T/null cell, primary systemic type Interdigitating dendritic cell sarcoma/tumor Hodgkin lymphoma (Hodgkin disease) Follicular dendritic cell sarcoma/tumor Nodular lymphocyte predominance Hodgkin lymphoma Dendritic cell sarcoma, not otherwise specified Classical Hodgkin lymphoma Nodular sclerosis Hodgkin lymphoma (Grades 1 and 2) Lymphocyte-rich classical Hodgkin lymphoma Mixed cellularity Hodgkin lymphoma nized, using morphologic, cytochemical, and im- Lymphocyte depletion Hodgkin lymphoma munophenotypic features. Recently, genetic fea- Only major categories are included. Subtypes and variants will be tures (cytogenetic and molecular genetic) as well as discussed in the WHO book (2) and in Tables 7–16. More common entities other features, such as previous therapy and a his- are underlined. a B and T/NK-cell neoplasms are grouped according to major clinical tory of myelodysplasia, have been shown to have a presentations (predominantly disseminated/leukemic, primary extran- significant impact on the clinical behavior of these odal, predominantly nodal). disorders, and these features do not always correlate perfectly with the FAB categories. Thus, a major focus of debate was how to integrate genetic and kemias, the classification of these disorders has not clinical features with morphology, cytochemistry, been recently updated. Thus, the discussion of and immunophenotype into a classification that these disorders generated considerable contro- can be used by pathologists and that will have clin- versy, and several subsequent meetings of pathol- ical relevance. A key issue, as with the lymphoid ogists and the clinical co-chairs occurred, during neoplasms, was to discriminate between disease en- which a consensus on the classification emerged. tities and prognostic factors. Some genetic abnor- The following summary includes both issues raised malities seem to define distinct diseases, whereas at the CAC meeting and resolutions achieved sub- others are prognostic factors within a given disease. sequently. Another issue debated was whether all diseases fit In the French-American-British (FAB) classifica- into one of the three major categories or additional tion, three main categories of myeloid neoplasms broad categories are needed. are recognized: acute myeloid leukemias, myelo- After discussion, it seemed that a paradigm similar dysplastic syndromes, and myeloproliferative dis- to that adopted for the Revised European-American orders (3). The blast count, lineage commitment, Classification of Lymphoid Neoplasms (REAL) can at and level of differentiation of the neoplastic cells least tentatively apply to the myeloid disorders; are the major determinants of the categories recog- namely, that a combination of morphology, immuno-

WHO Classification of Hematological Malignancies (N.L. Harris et al.) 195 phenotype, genetic features, and clinical features is monocytosis, and splenomegaly. It has been de- used to define distinct disease entities. The technol- bated whether this is really two diseases—one an ogy of genetic analysis is moving rapidly, and it is MDS and the other an MPD. However, studies have likely that advances in this field will necessitate revi- shown no differences in cytogenetic abnormalities, sions to any current classification in the near future. oncogene mutations, in vitro colony growth pat- The pathologists proposed four major groups of my- terns, or clinical outcome between the two types of eloid diseases: myeloproliferative diseases (MPD), CMML. It was the consensus of the meeting that myelodysplastic/myeloproliferative diseases (MD/ this is one disease. The committee concluded that it MPD), myelodysplastic syndromes (MDS), and acute fits better in the MPD than in the MDS category, but myeloid leukemias (AML). Within the category of after subsequent discussions, the pathologists rec- AML, four main groups are recognized: 1) AML with ommended that it be included in a separate cate- recurrent cytogenetic translocations, 2) AML with gory, with JMML, of disorders with both myelopro- myelodysplasia-related features, 3) therapy-related liferative and myelodysplastic features. AML and MDS, and 4) AML not otherwise specified What should be the nomenclature and category (NOS). for aCML? This disease was first recognized as a disease involving predominantly the neutrophil series, that lacked Ph1 or BCR/ABL translocation, Myeloproliferative Diseases which has dysplastic as well as proliferative fea- MPD are clonal stem cell disorders that are char- tures, often with multilineage dysplasia. The prog- acterized by “effective” hematopoiesis, resulting in nosis is significantly worse that that of Ph1ϩ CML. elevated peripheral blood levels of one or more cell It is clear that it is clinically, genetically, and mor- lines and hepatosplenomegaly; there is marrow hy- phologically distinct from Ph1ϩ CML, and the percellularity with maturation and without dyspla- name is therefore suboptimal, implying both a re- sia. Among the MPD, the prototype is Philadelphia lationship to Ph1ϩ CML and a chronic process. The ϩ ϩ chromosome (Ph1) [BCR/ABL ] chronic myelog- committee was unable to agree on another name enous leukemia (CML). The other accepted entities and believed that the term aCML could be retained, are polycythemia vera, idiopathic myelofibrosis, provided that a clear definition of the disease was and essential thrombocythemia. Controversies provided to prevent confusion. The pathologists within this group include the definitions and clas- recommended placing this disease with JMML and sification of juvenile myelomonocytic leukemia CMML in a category of myelodysplastic/myelopro- (also known as juvenile chronic myeloid leukemia liferative diseases. and juvenile chronic myelomonocytic leukemia), Should there be a separate category for cases that chronic myelomonocytic leukemia, and atypical are neither MDS nor MPD? For reasons mentioned CML (aCML). above, the pathologists recommended a fourth cat- Should juvenile myelomonocytic leukemia be a egory of myeloid neoplasms to contain those cases separate category? Should it be classified as MDS or that are inherently proliferative but show dysplastic MPD? The CAC accepted the conclusions of the features, including JMML, CMML, and aCML. It was international study group for pediatric MDS that the opinion of the clinicians present that such a juvenile myelomonocytic leukemia is a separate category was not desirable and that these diseases disorder, distinct from adult chronic myeloid or could be placed in the MPD category. However, the myelomonocytic leukemias. It has been proposed pathologists contended that these disorders have that the name juvenile myelomonocytic leukemia many features in common, including abnormalities (JMML) be adopted. The committee favored includ- of both granulocytic and monocytic lines and a ing it in the myeloproliferative disorders; however, relatively aggressive course, that distinguish them the pathologists recommended that a separate cat- from both the MDS and MPD categories and argued egory be formed to include this and other disorders for placing them together. that combine features of myeloproliferative and myelodysplastic syndromes. Summary Should chronic myelomonocytic leukemia (CMML) be divided into MDS and MPD types? 1. Should JMML be a separate category? Yes CMML has long been recognized as a disorder that 2. Should CMML be divided into MDS and MPD has features of both myelodysplastic and myelopro- types? No liferative syndromes. Nearly half of the patients 3. What should we call “atypical CML”? Atypical present with low or normal neutrophil counts, mul- CML tilineage marrow dysplasia, no organomegaly, and 4. Should there be a separate category for cases bone marrow morphology that resembles refractory that are neither MDS nor MPD? No consensus anemia with excess blasts (RAEB) but with mono- • Pathologists propose a category of MDS/ cytosis. Other patients have marked neutrophilia, MPD, to include JMML, CMML, and aCML.

196 Modern Pathology Acute Myeloid Leukemia and Myelodysplastic often show multilineage dysplasia or are pre- Syndromes ceded by a hypoproliferative state with multilin- What blast count should define AML? The FAB eage dysplasia, resembling MDS. Similar cytoge- standard has been 30% blasts. However, recent netic abnormalities are often seen in MDS not studies have indicated that patients with 20 to 30% associated with prior therapy, as well as in de blasts (classified as RAEB in transformation [RAEB- novo acute leukemias, particularly in older T]) have a prognosis similar to that of patients with adults. It has been suggested that all of these more than 30% blasts. Thus, there was a consensus disorders reflect similar genetic damage, which that the blast count for the diagnosis of AML should may be either environmental or iatrogenic. There be 20% and the category of RAEB-T should be was a consensus that the presence of multilin- dropped. eage dysplasia at the time of the diagnosis of Should cytogenetic/molecular categories of AML be acute leukemia, a history of myelodysplasia, and recognized as distinct diseases? Several specific cy- prior alkylating agent therapy all were adverse togenetic abnormalities in AML are associated with prognostic factors, which may reflect a common characteristic morphology and have distinctive pathogenesis. The committee concluded that clinical features. With the exception of promyelo- multilineage dysplasia, a history of MDS, and a cytic leukemia/M3 with t(15;17), these genetic ab- history of alkylating agent therapy should be in- normalities do not correlate precisely with FAB cat- cluded in the classification of AML. egories. The consensus of the CAC was that these The specific cytogenetic abnormalities com- categories should be recognized as distinct entities mon to MDS, alkylating agent–related AML, and within the classification. After discussion, the pa- poor-prognosis AML (3q-, -5, 5q-, -7, 7q-, ϩ8, ϩ9, thologists agreed that it would be possible to de- 11q-, 12p-, -18, -9, 20q-, ϩ21, t[1;7], t[2;11], com- velop morphologic criteria for these categories, plex karyotypes) likely reflect a common patho- which would permit them to be recognized, or at genesis of these lesions, distinct from that of least suspected, by pathologists, who should then other de novo AML. However, there was no con- suggest confirmation by genetic analysis. The spe- sensus on the role of these abnormalities in de- cific categories that will be defined are fining disease entities within the classification. Our understanding of this issue likely will im- 1. AML with t(8;21)(q22;q22), AML1(CBF␣)/ETO prove in the near future, necessitating a change 2. Acute promyelocytic leukemia (AML with t[15; in the major groupings. For the present, cytoge- 17][q22;q11–12] and variants, PML/RAR␣) netic abnormalities indicative of poor prognosis 3. AML with abnormal bone marrow eosinophils should be recognized as prognostic factors within (inv [16][p13q22] or t[16;16][p13;q22], CBF␤/ each category of AML. MYH11) Therapy with topoisomerase II inhibitors (epi- 4. AML with 11q23 (MLL) abnormalities podophyllotoxins and adriamycin) is also associated with secondary leukemias, which are often The specific morphologic features of these dis- myeloid but may be lymphoid. These typically orders will be described in the classification (2), show cytogenetic abnormalities associated with and these entities will be excluded from the FAB de novo AML—most commonly translocations categories used for cases that lack these abnor- involving 11q23 (MLL) but also occasionally t(8; malities. In addition, cases with these specific 21), inv (16), or t(15;17). These cases should also cytogenetic abnormalities with low blast counts, be recognized in the classification as distinct which might in the past have been diagnosed as from alkylating agent–related secondary leuke- MDS, will now be classified as AML. mias. Should multilineage dysplasia, prior MDS, Should refractory cytopenia with multilineage and/or prior therapy be included in classification dysplasia be a separate category? Myelodysplastic of AML? Severe multilineage dysplasia, defined as syndromes are clonal stem cell disorders charac- the presence of dysplastic features in the cells of terized by ineffective hematopoiesis, resulting two or more lines, has been shown to be associ- clinically in peripheral blood cytopenias; the ated with poor outcome in AML. Similarly, AML marrow is variably hypercellular, and patients arising in patients with a history of MDS also show poor responses to chemotherapy, with an have a poor prognosis. Therapy-related leuke- increased risk of progression to acute leukemia. mias secondary to alkylating agent therapy are The terms refractory anemia and refractory ane- clearly different from many de novo acute leukemia with ring sideroblasts were defined in the mias; they are associated with characteristic cy- FAB classification as having dysplasia largely re- togenetic abnormalities (3q-, -5, 5q-, -7, 7q-, ϩ8, stricted to the erythroid line. Recent studies have ϩ9, 11q-, 12p-, -18, -19, 20q-, ϩ21, t[1;7], t[2;11], shown that patients who have MDS with less than complex karyotypes) and a worse prognosis and 5% blasts but with significant dysplasia involving

WHO Classification of Hematological Malignancies (N.L. Harris et al.) 197 granulocytic and megakaryocytic lines have a plasms, T/NK-cell neoplasms, and Hodgkin dis- worse prognosis and are more likely to die of ease. The T- and B-cell neoplasms are stratified into marrow failure or progress to acute leukemia precursor, or lymphoblastic, neoplasms (acute lym- (similar to RAEB) than those lacking these fea- phoblastic leukemia and lymphoblastic lymphoma) tures. Thus, the committee agreed that a separate and mature (“peripheral”) B- and T-cell neoplasms. category is needed for these cases. Multilineage The mature B- and T-cell neoplasms are informally dysplasia is defined as the presence of dysplastic grouped according to their major clinical presenta- features in the cells of two or more lines. Refrac- tions: predominantly disseminated/leukemic, pri- tory anemia (with or without ring sideroblasts) mary extranodal, and predominantly nodal dis- will continue to be defined as a disorder involving eases. The pathologists sought input from the the erythroid line only. MDS will exclude cases of clinicians on these changes and on some issues that low blast-count leukemias that show one of the remain controversial or problematic, such as grad- AML-type cytogenetic abnormalities—t(8;21), inv ing of follicular lymphoma, how to define “Burkitt- (16), or t(15;17). Because of the distinctive mor- like” lymphoma, subclassification of large B-cell phologic and clinical features of the 5q- syn- lymphomas and mature T-cell lymphomas, and the drome, it was agreed by the pathologists that this desirability of clinical groupings of the non- should be a separate category within MDS. Hodgkin lymphomas.

Summary Precursor Neoplasms 1. What blast count should define AML? 20% Should the FAB terms (L1, 2, 3) be retained? There • Eliminate RAEB-T was a consensus that these terms are no longer 2. Should cytogenetic/molecular categories be relevant, because L1 and L2 morphology do not recognized as distinct diseases? Yes predict immunophenotype, genetic abnormalities, • t(8;21)(q22;q22), AML1(CBF␣)/ETO or clinical behavior. L3 is generally equivalent to • Acute promyelocytic leukemia t(15;17)(q22; Burkitt lymphoma in leukemic phase and should be q11–12), PML/RAR␣ and variants diagnosed as such. • Acute myeloid leukemia with abnormal bone Are lymphoblastic leukemias and lymphoblastic marrow eosinophils (inv [16][p13q22] and lymphomas a single disease with different presenta- variants, CBF␤/MYH11) tions? There was a consensus that the precursor • 11q23, MLL abnormalities neoplasms presenting as solid tumors and those 3. Should severe multilineage dysplasia, prior presenting with marrow and blood involvement are therapy, and/or prior MDS be included in biologically the same disease but with different classification of AML? Yes clinical presentations The presence of bone mar- 4. Should MDS with multilineage dysplasia be a row and peripheral blood involvement are princi- separate category? Yes pally prognostic factors/staging issues, not classification issues, although the biologic basis for the different clinical presentations is not fully under- LYMPHOID NEOPLASMS stood. Most precursor lymphoid neoplasms present The proposed WHO classification of lymphoid as leukemia, and thus it was agreed that the classi- neoplasms adopts the Revised European-American fication should retain the term acute lymphoblastic Classification of Lymphoid Neoplasms (REAL), pro- leukemia, for the leukemic phase of precursor neo- posed by the International Lymphoma Study plasms of T and B types (Table 6). Group. This classification is based on the premise Should genetic abnormalities be included in the that a classification should attempt to define dis- classification? Genetic abnormalities are important tinct disease entities, using all available informa- prognostic factors within precursor B lymphoblas- tion, including morphology, immunophenotype, tic neoplasms (t[9;22][q34;q11], BCR/ABL; 11q23, genetic features, and clinical features. There is no MLL; t[1;19][q23;p13], E2A/PBX1; t[12;21][p12;q22]; ␣ one gold standard, and the importance of various ETV/CBF ]), and pathologists who undertake to di- criteria for both definition and diagnosis differs among different diseases. On the basis of experience with using this classification for several years TABLE 6. Acute Lymphoid Leukemias and on input from the committees, several changes Precursor B-cell acute lymphoblastic leukemia (cytogenetic subgroups) were proposed for the WHO version. These include t(9;22)(a34;q11); BCR/ABL t(v;11q23); MLL rearranged some changes in nomenclature, splitting some cat- t(1;19)(q23;p13) E2A/PBX1 egories that were believed to be heterogeneous, and t(12;21)(p12;q22) ETV/CBF␣ adopting some “provisional” entities as “real.” The Precursor T-cell acute lymphoblastic leukemia proposed classification recognizes B-cell neo- Burkitt cell leukemia

198 Modern Pathology agnose these neoplasms should be familiar with the follicle center derivation, such as BCL2 rearrange- types and significance of genetic abnormalities that ment or CD10 expression. can be seen. The genetic analysis should be part of Should follicular lymphoma be graded by the (or an addendum to) the pathology report when- number of large cells? The following points were ever feasible. made. First, follicular lymphoma of Grade 1 (follic- Summary ular small cleaved) and Grade 2 (follicular mixed) are more closely related to each other than to Grade 1. Should the FAB terms (L1, 2, 3) be retained? 3 (follicular large cell [FLC]), because in sequential No biopsies, transitions are seen from Grade 1 (follic- 2. Are acute lymphoblastic leukemias and lym- ular small cleaved) to Grade 2 (follicular mixed) and phoblastic lymphomas a single disease with vice versa but rarely from Grade 1 to Grade 3 (FLC). different clinical presentations ? Yes Second, patients with Grade 3 (FLC) tend to have • Retain the term leukemia for all precursor T earlier relapses (worse freedom from relapse) than and B types Grades 1 and 2 but similar overall survival, and this 3. Should cytogenetics be included in classifica- inferior freedom from relapse may be obliterated by tion? Yes adriamycin-containing therapy. Third, Grade 3 fol- • As prognostic factors within each subtype licular lymphoma is not the same disease as diffuse • t(9;22)(q34;q11), BCR/ABL; 11q23, MLL; t(1; large B-cell lymphoma (DLBCL), because it has a 19)(q23;p13), E2A/PBX1; t(12;21)(p12;q22), higher relapse rate although slightly better overall ␣ ETV/CBF survival. Finally, pathologists discriminate poorly between follicular lymphoma of Grades 1 and 2 but Mature B and T/NK Neoplasms may be better able to discriminate between these As for the precursor neoplasms, the proposed and Grade 3 cases. Several studies suggest that the classification considers lymphomas and lymphoid “Berard” criteria (3) for the diagnosis of Grade 3 leukemias of the same cell type as one disease with follicular lymphoma (more than 15 centroblasts/ different clinical presentations or stages. For the high power field [HPF]) may best define the group mature B and T/NK neoplasms, this question is of cases with a potential for early relapses that may primarily relevant to B-cell chronic lymphocytic be prevented by adriamycin-containing chemo- leukemia and B-cell small lymphocytic lymphoma. therapy. There was no consensus on whether this is Although patients in some locations may be seen by warranted as initial therapy for these patients. It different physicians based on their presentation was also noted that other factors than histologic (e.g., those presenting with peripheral blood in- grade affect outcome in patients with follicular lym- volvement leukemias being seen by hematologists phoma, including clinical features summarized in and those presenting with tissue involvement lym- the International Prognostic Index and potential phomas by oncologists), there was a consensus that biologic markers such as BCL2 expression and p53 they are biologically the same disease (Table 7). mutations. In summary, there was a consensus that follicular Follicular Lymphoma lymphoma should be graded, at least into two Should the nomenclature be changed to follicular grades, with what is currently recognized as Grade lymphoma? The WHO committee proposed to 3 (FLC) being discriminated from lower grade cases. change the nomenclature from “follicle center lym- Although there are minor differences in natural his- phoma” to “follicular lymphoma.” The CAC over- tory and response to treatment between Grades 1 whelmingly approved this proposal. For the rare and 2 follicular lymphoma, there was a consensus case of purely diffuse lymphoma that seems to be of that these did not mandate different approaches to follicle center origin (predominance of centrocytes, treatment and thus were not of great clinical im- rare centroblasts, BCL2 rearranged), the term folli- portance. Nonetheless, there was concern that cle center lymphoma, diffuse will be retained as a changing the nomenclature would be potentially separate category. This diagnosis should be made confusing and that a three-grade system should be only when both small and large cells are B cells and retained. The pathologists were encouraged to de- preferably with demonstration of some indicator of fine clinically relevant and reproducible criteria for such grading. After discussion, the pathologists concluded that because only the Berard cell- TABLE 7. B-Cell Neoplasms, Predominantly Disseminated/Leukemic Types: Variants counting method (Table 8) has been repeatedly tested in the literature, it should be recommended B-cell chronic lymphocytic leukemia/small lymphocytic lymphoma Variant: with monoclonal gammopathy/plasmacytoid differentiation for use (Grade 1: 0–5 centroblasts/HPF; Grade 2: Hairy cell leukemia 6–15 centroblasts/HPF; Grade 3: more than 15 cen- Variant: hairy cell leukemia variant troblasts/HPF. Ten to 20 HPFs, within different fol-

WHO Classification of Hematological Malignancies (N.L. Harris et al.) 199 TABLE 8. Follicular and Mantle Cell Lymphomas: (Ͼ75% follicular), follicular and diffuse (25 Grading and Variants to 75% follicular), predominantly diffuse Follicular lymphoma (Ͻ25% follicular). Grades: • Grade 1: 0–5 centroblasts/HPF Areas of DLBCL should be classified sepa- Grade 2: 6–15 centroblasts/HPF rately. Example of suggested terminology: Grade 3: Ͼ15 centroblasts/HPF follicular lymphoma, Grade 3/3 (75%), with 3a: Ͼ15 centroblasts, but centrocytes are still present 3b: Centroblasts form solid sheets with no residual centrocytes DLBCL (25%). Variants: Cutaneous follicle center lymphoma Marginal Zone Lymphomas Diffuse follicle center lymphoma Grade 1: 0–5 centroblasts/HPF Should the term extranodal marginal zone B-cell Grade 2: 6–15 centroblasts/HPF lymphoma of mucosa-associated lymphoid tissue Mantle cell lymphoma (MALT) or MALT-type lymphoma be applied only to Variant: blastoid a lymphoma composed mostly of small cells? What HPF, high power field. should be the terminology for large-cell lymphoma in a MALT site? The term high-grade MALT lymphoma, which is used by some pathologists to de- licles are counted; these are representative follicles, note either transformation of a low-grade MALT not selected for those with the most numerous large lymphoma or any large B-cell lymphoma in a MALT cells) (4). site, is confusing to clinicians, who have come to Should diffuse areas be reported? Several oncolo- regard the term MALT lymphoma as synonymous gists believed that diffuse areas in all grades of with a lesion that may respond to antibiotic therapy follicular lymphoma do seem to have an impact on for eradication of Helicobacter pylori. Because ex- prognosis. There was a consensus that diffuse areas perience indicates that patients with a component should be reported and quantified according to the of large-cell lymphoma may not respond to antibi- recommendations of the REAL classification: pre- otic therapy, the oncologists were concerned that dominantly follicular (Ͼ75% follicular), follicular use of this term may result in undertreatment in and diffuse (25 to 75% follicular), and predomi- cases of extranodal large-cell lymphoma. Further- nantly diffuse (Ͻ25% follicular). However, it is not more, recent data show that the types of cytoge- clear what the implications of these features for netic abnormalities seen in low-grade MALT lym- treatment would be. In Grade 3 follicular lymphomas differ from those seen in primary large-cell phoma, diffuse areas represent areas of DLBCL and lymphoma of the stomach, raising the question of should be reported as such (e.g., “follicular lym- whether these primary lymphomas are really re- phoma, Grade 3/3 [75%] with diffuse large B-cell lated to low-grade MALT lymphomas. Therefore, lymphoma [25%],” not “follicular lymphoma, Grade the oncologists preferred that the term MALT lym- 3, follicular and diffuse.”) The presence of DLBCL in phoma be used only for the low-grade lymphoma any follicular lymphoma will dictate more aggres- originally described as “low-grade B-cell lymphoma sive therapy. of MALT.” Areas of large-cell lymphoma, if present, should be separately diagnosed as “DLBCL.” Pri- Summary mary large-cell lymphomas of MALT sites should be 1. Change nomenclature from “follicle center diagnosed as “DLBCL,” not as “high-grade MALT lymphoma” to “follicular lymphoma”? Yes lymphoma.” 2. Should it be graded by the number of large Should marginal zone/MALT lymphoma be cells? Yes graded by the proportion of large cells? The issue of 3. Are two grades adequate for clinical practice? grading MALT lymphoma has not been extensively Yes studied. Several early reports suggested that cases • But three grades will be used to avoid con- with up to 25% large cells did not have a worse fusion. prognosis than cases with fewer large cells. How- 4. What should be method of grading? No con- ever, a recent report of patients treated primarily sensus with antibiotics found that the presence of in- • Pathologists recommend cell-counting creased transformed cells (5 to 10% with clusters of method. fewer than 20 cells) conferred a slight but signifi- • Grade 1 (1–5 centroblasts/HPF); Grade 2 cantly worse prognosis compared with cases with (6–15 centroblasts/HPF); Grade 3 (Ͼ15 cen- fewer than 5% large cells. Cases with high-grade troblasts/HPF) areas consisting of sheets of blasts (Ͼ20 cells) be- 5. Should diffuse areas be reported? Yes haved similarly to large-cell lymphoma with no 6. How should they be quantified? No consensus low-grade component. In addition, it was reported • Pathologists recommended criteria sug- at the meeting that the International Non- gested in REAL classification: follicular Hodgkin’s Lymphoma classification project indi-

200 Modern Pathology cated that the presence of more than 5% large cells B-Cell Chronic Lymphocytic Leukemia/Small in an extranodal marginal zone lymphoma con- Lymphocytic Lymphoma ferred a worse prognosis, as did areas of DLBCL. Are B-cell chronic lymphocytic leukemia (CLL) The consensus of the committee was that the data and small lymphocytic lymphoma (SLL) one disease available raise the concern that increased large cells at different stages? As for the precursor neoplasms may be of prognostic importance in MALT lym- and Burkitt lymphoma, the committee agreed with phoma and warrant further study. The WHO clas- the pathologists that B-CLL and SLL are one disease sification should specify criteria for grading so that at different stages, not two separate entities, its significance can be tested in future clinical stud- and should be listed together in the classification ies. In cases of marginal zone B-cell lymphoma (Table 7). (low-grade MALT lymphoma) with coexisting Are cases of B-CLL with plasmacytoid differentia- DLBCL, a separate diagnosis of DLBCL should be tion (lymphoplasmacytoid immunocytoma in the made. The principle is therefore similar to that for Kiel Classification) a different disease from typical follicular lymphoma: the tumors are graded accord- CLL? Data from several groups using the Kiel Clas- ing to the number of large cells, but when confluent sification suggest that plasmacytoid differentiation areas of large cells are present, this indicates trans- may be an adverse prognostic factor in B-CLL; the formation to DLBCL. committee believed that the available data do not Are marginal zone lymphomas of nodal and support calling it a different disease and that fur- splenic type “real”? There was a consensus that re- ther study is needed to determine whether plasma- cent data support the recognition that two other cytoid differentiation is an adverse prognostic fac- types of lymphoma called “marginal zone lympho- tor in CLL. Therefore, recognition of this feature is mas” are distinct both from MALT lymphoma and not required for diagnosis for clinical purposes, but from each other. Splenic marginal zone lymphoma criteria for diagnosing plasmacytoid differentiation seems to be the tissue counterpart of splenic lym- should be agreed on if possible for future studies. phoma with villous lymphocytes. Patients typically Summary are older adults with bone marrow and blood involvement and a very indolent clinical course. 1. Are B-CLL and SLL one disease at different Nodal marginal zone lymphoma (which often has a stages? Yes prominent monocytoid B-cell component) must be 2. Is plasmacytoid differentiation an indication distinguished both from MALT lymphoma with of a different disease? No lymph node involvement and from other lympho- 3. Is plasmacytoid differentiation a prognostic mas (particularly follicular and mantle cell lym- factor? Research question phoma) with a marginal zone pattern or a compo- Mantle Cell Lymphoma nent of monocytoid B cells. Nodal marginal zone lymphoma seems to have a high rate of early re- Should mantle cell lymphoma be subclassified/ lapse and overall survival similar to or slightly graded for clinical purposes? A number of studies worse than that of follicular lymphoma. have found morphologic heterogeneity in mantle cell lymphoma (MCL) in both pattern and cytology Summary and have suggested that some features may predict outcome. For example, cases with a mantle zone 1. Should the term extranodal marginal zone pattern have been less aggressive in some studies B-cell lymphoma of MALT,orMALT-type lym- but not in others, and cases with blastic or blastoid phoma, be applied only to a lymphoma com- morphology have had a worse prognosis in some posed mostly of small cells and not to large- reports. It was the consensus of the committee that cell lymphoma in a MALT site? Yes because no effective therapy exists for any type of 2. Should the term high-grade MALT lymphoma MCL, stratification by morphologic features is not be used? No required for clinical diagnostic purposes at this • Suggested terminology: DLBCL (Ϯareas of time. However, the different cytologic types and marginal zone/MALT-type lymphoma) patterns will be included in the text of the classifi- 3. Should extranodal marginal zone B-cell lym- cation (2) so that variant cases will be recognized as phoma of MALT type be further graded/strat- MCL for diagnosis and graded similarly for research ified based on number of large cells? Research studies (Table 8). question Summary • Criteria should be given so that additional studies can be done. 1. Should MCL be subclassified/graded by cytol- 4. Are nodal and splenic marginal zone lympho- ogy for clinical purposes? No mas distinct diseases that should be recog- 2. Should MCL be subclassified/graded by pat- nized and defined in the classification? Yes tern for clinical purposes? No

WHO Classification of Hematological Malignancies (N.L. Harris et al.) 201 • Different cytologic types and patterns nonreproducible category, with only approximately should be included so that they will be rec- 50% agreement among the pathologists; the major ognized as MCL for diagnosis and graded areas of overlap were DLBCL and Burkitt lym- similarly for research. phoma. The oncologists urged that the category of Burkitt-like lymphoma be reserved for tumors that Large B-cell Lymphoma and Burkitt-Like should be treated “like Burkitt lymphoma”—that is, Lymphoma very high-grade tumors. The committee concluded Should morphologic subclassification of DLBCL that Burkitt-like lymphoma should be listed as a be required? There was a consensus of the CAC that morphologic variant of Burkitt lymphoma in the neither biologic nor clinical data at present support WHO classification. The term atypical Burkitt lym- a requirement for subclassification of DLBCL ac- phoma was proposed for this variant; however, the cording to the criteria of the Working Formulation Steering Committee subsequently decided that the or the Kiel Classification. Data from the Kiel group term Burkitt-like was preferable, because the rela- suggest that immunoblastic lymphoma as defined tionship to Burkitt lymphoma is not known in all in the updated Kiel Classification (Ͼ90% immuno- cases. Thus, the category of Burkitt lymphoma will blasts) has a worse prognosis than centroblastic include classic Burkitt lymphoma and a variant, lymphoma. Other data suggest that staining for Burkitt-like lymphoma. In addition, three subcate- bcl-6 (centroblastic) and syndecan-1/CD138 (im- gories—endemic, nonendemic, and immunodefi- munoblastic) or evidence of BCL6 rearrangement ciency associated—were proposed to reflect the (centroblastic) may help to discriminate between major clinical and genetic subtypes of this disease. them. Nonetheless, neither reliable pathologic or At present, there are no readily available immu- biologic criteria for subclassification nor distinctive nophenotypic criteria that can be used in this dif- therapies that can be recommended for clinical ferential diagnosis. However, participants observed practice are available at this time. For these rea- that probably both the morphology and the biology sons, the committee believed that these categories of Burkitt lymphoma are defined by the presence of should remain optional at this time. However, there cMYC rearrangement and overexpression, which was agreement that the pathologists should de- results in all cells being perpetually in cycle. The velop criteria for subclassification so that these cat- gold standard for the diagnosis of Burkitt lym- egories can be tested in future clinical studies (Ta- phoma should be the presence of the t(8;14)(q24; ble 9). q32) and its variants or cMYC rearrangement. Cy- Should “Burkitt-like” or “non-Burkitt” lymphoma togenetic analysis is recommended in all leukemic be a subtype of DLBCL, a subtype of Burkitt lym- cases. If cytogenetic or Southern blot analysis is not phoma, or a distinct category? What should be de- available in solid tumors, it seems likely that the fining criteria? The pathologists proposed to define most reasonable surrogate for cMYC rearrange- Burkitt-like lymphoma as a subtype of large B-cell ment is proliferation fraction. Therefore, it was sug- lymphoma. However, there was a clear consensus gested that cases in which cytogenetic analysis is among the oncologists that this would be a mistake. not available should not be diagnosed as Burkitt There are abundant data indicating that in children, lymphoma or Burkitt-like lymphoma without a cases classified as Burkitt-like (or non-Burkitt) be- Ki-67 fraction close to 100%. Thus, the definition of have identically to Burkitt lymphoma and would be Burkitt-like lymphoma is a lymphoma that mor- undertreated if treated as large B-cell lymphoma. In phologically resembles Burkitt lymphoma but has adults, the biology of cases classified as Burkitt-like more pleomorphism or large cells than classical is less clear, but this may reflect the heterogeneity Burkitt lymphoma and has a proliferation fraction of the diagnostic criteria. In the International Non- of more than 99%. Hodgkin’s Lymphoma study, Burkitt-like was a Do we need separate categories for clinical subtypes of DLBCL? There are multiple distinct clinical presentations of DLBCL, several of which have TABLE 9. Diffuse Large B-Cell Lymphoma: Morphologic unique clinical behavior. These include mediasti- Variants and Subtypes nal/thymic large B-cell lymphoma, primary central Diffuse large B-cell lymphoma, morphologic variants nervous system (CNS) lymphoma, and primary ef- Centroblastic fusion lymphoma. Of particular concern to pathol- Immunoblastic T-cell/histiocyte rich ogists is the category of cutaneous B-cell lym- Lymphomatoid granulomatosis type phoma, most of which have a very indolent clinical Anaplastic large B-cell course. One category—marginal zone/MALT lym- Plasmablastic Diffuse large B-cell lymphoma, subtypes phoma—is easily recognized by pathologists as a Mediastinal (thymic) large B-cell lymphoma low-grade lymphoma. However, the other major Primary effusion lymphoma category, called cutaneous follicle center lymphoma Intravascular large B-cell lymphoma in the recently proposed European Organization for

202 Modern Pathology Research and Treatment of Cancer (EORTC) classi- in patients with HIV-positive status. The recently fication, has a range of morphology, from a clearly described primary effusion lymphoma, which was low-grade lesion resembling nodal follicular lym- initially thought to be unique to patients with HIV- phoma to a diffuse proliferation with numerous positive status, has been reported in patients with large cells that may be called DLBCL by patholo- HIV-negative status as well. T-cell lymphomas in gists. This type of lymphoma, which typically is patients with HIV-positive status also do not seem localized to the head and trunk, responds well to to be distinctive. A recently described plasmablastic local therapy (excision or radiation), and typically lymphoma is distinctive, and its relationship to my- does not disseminate to lymph nodes, composed eloma remains to be determined. 70% of cutaneous B-cell lymphomas in the EORTC The polymorphic posttransplant lymphoprolif- study. There is concern that if its distinctive histo- erative disorders (PTLD) seem to be a unique form logic and clinical features are not recognized by of lymphoproliferation that does not occur in im- both pathologists and oncologists, these patients munologically normal individuals. It was suggested will be overtreated with aggressive chemotherapy. that EBNA-2 expression in these lesions indicates The consensus of the committee was that sepa- that the proliferation is EBV driven and may re- rate classifications of lymphomas at specific extran- spond to reduced immunosuppression. odal sites were not needed for clinical purposes. However, the site of involvement should be clearly In summary, the committee suggested that a stated in the pathology report, and oncologists are separate classification was not needed for obliged to understand the distinctive clinical fea- immunodeficiency-associated lymphomas but that tures of lymphomas at various sites. Distinct enti- the specific types of lymphomas that occur in im- ties such as primary mediastinal (thymic) B-cell munodeficiency states and their distinctive features lymphoma, primary effusion lymphoma, and intra- in these conditions should be indicated both in the vascular lymphoma will be described in the text as text and in a table. In addition, the pathologists subtypes of DLBCL (Table 9). The committee rec- believed that a separate classification of PTLD ommended that the distinctive clinical features of would be useful, because of their distinctive bio- B-cell lymphomas in the skin be indicated in the logic and clinical features (Table 3). text describing each lymphoma subtype. Summary Summary Do we need a separate classification for lympho- 1. Should morphologic subclassification of DLBCL be required? No mas in immunodeficiency states? No • • Criteria for subclassification should be stan- Note the frequency of specific types in immu- dardized for future studies. nodeficiency states. • 2. Should the category of “Burkitt-like” be a sub- PTLD are distinctive and need a separate clas- type of large B-cell lymphoma? No sification. • Burkitt-like lymphoma will be considered a • EBV status may be important in determining variant of Burkitt lymphoma. prognosis/treatment. • Major criteria include 1) morphology inter- mediate between Burkitt lymphoma and Are Clinical Syndromes Integral to the Definition large-cell lymphoma, 2) t(8;14)(q24;q32) of T/NK-Cell Neoplasms? and variants, cMYC rearrangement, or 3) Many distinct T- and/or NK-cell diseases have a proliferation fraction (Ki-67) more than range of cytologic composition (small to large to 99%. anaplastic). Immunophenotypic variation exists 3. Do we need separate categories for clinical within disease entities, and many antigens are subtypes of DLBCL? No • Location should be indicated in report. shared by different diseases. Specific cytogenetic features are not defined for most entities, and even Lymphomas in Immunodeficiency States: Do We T-cell receptor types (␣␤ versus ␥␦)orTversus NK Need a Separate Classification? lineage is not sufficient to define distinct disease Most lymphomas that occur in immunodefi- entities. To a greater extent than is appreciated for ciency states are also seen in nonimmunosup- B-cell neoplasms, it seems that clinical syndromes, pressed patients but have some distinctive features and particularly location (nodal versus extranodal in patients immunodeficiency. For example, in pa- and specific extranodal sites), are important in de- tients with HIV-positive status, primary CNS lym- termining the biologic behavior of the disease. The phoma is always Epstein-Barr virus (EBV) positive, committee agreed that clinical syndromes seem to in contrast to sporadic CNS lymphoma. Hodgkin be integral to the definition of T- and NK-cell neo- disease is more aggressive and always EBV positive plasms.

WHO Classification of Hematological Malignancies (N.L. Harris et al.) 203 Should Peripheral T-Cell Lymphoma, Unspecified, TABLE 10. Burkitt Lymphoma: Morphologic Variants and Subtypes Be Subclassified (According to the Kiel Classification) for Clinical Purposes? Burkitt lymphoma, morphologic variants Burkitt-like On the basis of the available data, there seems to With plasmacytoid differentiation (AIDS associated) be no immediate justification or clear criteria for Burkitt lymphoma, subtypes (clinical and genetic) Endemic recognizing cytologic subtypes within this broad Sporadic category. However, given the marked differences in Immunodeficiency associated clinical behavior between primary extranodal T/NK-cell lymphomas and primary nodal lymphomas, it is likely to be clinically relevant to subdivide cutaneous lymphoproliferative disease on the pa- the “unspecified” category into nodal and extra- thology reports, with the understanding that clini- nodal types. Both pathologists and oncologists will cal criteria must be added to determine whether the need to continue to address this area in further patient has a locally progressive disease that re- studies (Tables 13–15). quires treatment (ALCL) or a relapsing condition that needs no treatment (lymphomatoid papulosis). Summary What is the gold standard for defining ALCL? 1. Are clinical syndromes integral to the defini- Given the recent availability of an antibody to the tion of peripheral T/NK-cell neoplasms? Yes ALK protein, which is highly associated with the 2. Is cytologic subclassification of peripheral t(2;5)(p23;q35), the question raised was whether T-cell lymphoma required for clinical pur- this can be used as the defining criterion for ALCL. poses? No Clinically, cases with the t(2;5) and/or ALK positiv- ity seem to represent a homogeneous group with a Anaplastic Large Cell Lymphoma relatively good prognosis. However, others ob- Should cutaneous and systemic anaplastic large served that experience with ALK antibodies is lim- cell lymphoma (ALCL) be considered one disease or ited and they are only now becoming commercially two? What should be the terminology for the cuta- available. In addition, there are cases with typical neous type? There is evidence that most cases of morphology and immunophenotype that are ALK ALCL of T-cell type presenting with disease local- or t(2;5) negative. The committee concluded that a ized to the skin are a different disease from sys- single gold standard for the diagnosis of ALCL does temic ALCL: the clinical course is indolent, they not exist; the diagnosis requires both morphology lack the t(2;5)(p23;q35) and are ALK protein nega- and immunophenotype, and at least at present, tive, and seem to form a spectrum with lymphoma- restricting the diagnosis to ALK-positive cases does toid papulosis. Although some members of the not seem to be justified. It was suggested that ALK committee believed that the clinical course was not staining be done in all cases to the extent possible predictably indolent, there was general agreement and that cases be designated as ALCL, ALK positive, that at least for the purposes of further study, cu- or ALK negative, at least for research purposes. In taneous and systemic ALCL should be considered addition, pathologists need to be aware of the distinct categories. There was significant concern, broad morphologic spectrum of ALCL. however, about the proposed term primary CD30ϩ cutaneous lymphoproliferative disorder—a term Summary that includes lymphomatoid papulosis, cutaneous 1. Is cutaneous ALCL different from systemic ALCL, and CD30ϩ cutaneous T-cell lymphomas ALCL? Probably that do not have typical “anaplastic” morphology. • Distinction between them is not always Oncologists believed that including lymphomatoid straightforward, and cutaneous type is not papulosis in a classification of lymphomas would always indolent. imply to patients and insurers that this is a malignancy, whereas it typically has a benign clinical course. TABLE 11. Plasma Cell Disorders: Subtypes and In conclusion, the committee agreed that the en- Variants tity, primary cutaneous ALCL, should be included Monoclonal gammopathy of undetermined significance Plasma cell myeloma variants in the list of neoplasms and that a discussion of Indolent myeloma CD30ϩ cutaneous lymphoproliferative diseases Smoldering myeloma should be included in the text with a discussion of Osteosclerotic myeloma (POEMS syndrome) Plasma cell leukemia lymphomatoid papulosis and borderline lesions. Nonsecretory myeloma Because of the difficulty in predicting by morphol- Plasmacytoma variants ogy alone which disease the patient has, patholo- Solitary plasmacytoma of bone gists will often be forced to use the term CD30ϩ Extramedullary plasmacytoma

204 Modern Pathology TABLE 12. Immunosecretory Disorders (Clinical Manifestations of Diverse Lymphoid Neoplasms)

Clinical Syndrome Underlying Neoplasm Waldenstrom’s macroglobulinemia Lymphoplasmacytic lymphoma Heavy chain diseases (HCD) gamma HCD Lymphoplasmacytic lymphoma alpha HCD Extranodal marginal zone lymphoma (immunoproliferative small intestinal disorder) Mu HCD B-cell chronic lymphocytic leukemia Immunoglobulin deposition diseases Systemic light chain disease Plasma cell myeloma, monoclonal gammopathy Primary amyloidosis Plasma cell myeloma, monoclonal gammopathy

2. Should lymphomatoid papulosis appear in the Is anaplastic large cell lymphoma, Hodgkin like, list of lymphoid neoplasms? No real? The pathologists proposed to drop this provi- • It should be discussed in the text along with sional category from the REAL classification, believ- borderline cases. ing that there is probably no true biologic border- 3. Is there a gold standard for the diagnosis of line between Hodgkin disease (in most cases a ALCL? Not yet B-cell process) and ALCL (in most cases a T-cell • The morphologic spectrum of ALCL needs process). Some cases of ALCL may have a nodular to be better understood by pathologists. growth pattern and areas of fibrosis and thus re- • Cases should be listed as ALK positive or semble Hodgkin disease of nodular sclerosis type. ALK negative for research. Some cases of nodular sclerosis Hodgkin disease may have increased numbers of malignant cells and Hodgkin Disease therefore resemble ALCL. However, this resem- Should grading of nodular sclerosis be required for blance does not indicate a biologic relationship. clinical use? Data on the clinical impact of grading Pathologists should strive to resolve morphologi- nodular sclerosis Hodgkin disease according to the cally difficult cases by immunophenotyping and, if British National Lymphoma Investigation criteria necessary, molecular genetic studies. In a case that (Grade 1 ϭ few RS cells; Grade 2 ϭ many RS cells) is morphologically on the borderline between have shown conflicting results, with some studies Hodgkin disease and ALCL, expression of CD15 showing that Grade 2 cases are associated with a with or without B-cell antigens favors Hodgkin dis- worse outcome and others showing no difference in ease, whereas absence of CD15 and expression of outcome. The committee recommended that grad- T-cell antigens or ALK protein favor ALCL. Detec- ing not be required for clinical purposes in routine tion of T-cell receptor gene or NPM/ALK rearrange- diagnosis but that the classification include clear ment would confirm T-cell lymphoma, and absence criteria so that this question can be tested in future of rearrangements would favor Hodgkin disease. studies. Cases that cannot be resolved by a combination of Nomenclature: Hodgkin disease or Hodgkin lym- morphology, immunophenotype, and genetic stud- phoma? Because it is now clear that Hodgkin dis- ies should be considered unclassifiable. Clinical ease is a clonal proliferation of (in most cases) B judgment should be used to determine whether to cells and therefore qualifies as a lymphoma, the rebiopsy or to treat with a regimen that would be pathologists proposed that the name be changed to suitable for both Hodgkin disease and ALCL. Hodgkin lymphoma. Opinion of the committee was divided on this score; some believed that patients Summary become confused as to whether they have a lymphoma or not when the term disease is used, and 1. Should grading of nodular sclerosis Hodgkin others stood on tradition and resisted unnecessary disease be required for clinical use? No change. No consensus was reached. • Is lymphocyte-rich classical Hodgkin disease a Criteria need to be clearly defined for future “real” subtype? Very few clinical data exist on this studies. subtype, proposed as “provisional” in the REAL 2. Should lymphocyte-rich classical Hodgkin dis- classification. The committee agreed that it was ease be a separate category? Yes • important to separate these cases from nodular Clinical features need further study. lymphocyte predominance Hodgkin disease for 3. Is ALCL-Hodgkin disease-like a real entity? No clinical purposes and that it would be valuable to • Pathologists should use immunophenotyp- separate them from other types of classical HD for ing and molecular genetic techniques to clinical research purposes. classify morphologically borderline cases as

WHO Classification of Hematological Malignancies (N.L. Harris et al.) 205 either Hodgkin disease or ALCL; unresolved TABLE 14. Peripheral T-Cell Neoplasms, Primary Extranodal Types: Variants and Subtypes cases should be called unclassifiable. 4. Should we change the name from Hodgkin Mycosis fungoides variants Pagetoid reticulosis disease to lymphoma? No consensus Mycosis fundoides–associated follicular mucinosis • Proposal: allow both (Hodgkin disease/ Granulomatous slack skin disease Hodgkin lymphoma) Primary cutaneous CD-30 positive T-cell lymphoproliferative disorders Lymphomatoid papulosis (type A and B) Primary cutaneous anaplastic large cell lymphoma Borderline lesions Clinical Groupings of B- and T/NK-Cell Lymphomas

Are clinical groupings of B- and T/NK-cell lym- TABLE 15. Peripheral T-Cell Neoplasms, Predominantly phomas useful for clinical practice? The committee Nodal Types: Variants concluded that grouping the B- and T/NK-cell neo- Peripheral T-cell lymphoma (not otherwise categorized), variants plasms into prognostic categories would serve no Lymphoepithelioid (Lennert’s) T-zone clear purpose and could hamper understanding of Anaplastic large cell lymphoma T/null cell type, variants the specific features of some of the diseases. There Lymphohistiocytic are no groups of diseases that require identical Small cell treatment, and if treatment must be individualized to a specific disease, grouping serves no purpose TABLE 16. Proposed Categories of Unclassifiable and may be misleading. The entities listed in the Hematologic Malignancies classification are clearly defined and clinically rel- Hematologic malignancy, unclassifiable evant, and it is necessary for oncologists and pa- Myeloid neoplasm, unclassifiable thologists dealing with these diseases to under- Myeloproliferative disease, unclassifiable Myelodysplastic syndrome, unclassifiable stand each of them. Acute myeloid leukemia, unclassifiable Is a shorter list of diseases necessary for clinicians? Lymphoid neoplasm/lymphoma, unclassifiable The committee also discussed whether a shorter list B-cell lymphoma, unclassifiable T-cell lymphoma, unclassifiable of common diseases should be prepared for clinical Hodgkin disease, unclassifiable use. There was a clear consensus that the complete Histiocytic neoplasm, unclassifiable list of neoplasms should have more common entities highlighted, to draw the attention of nonexperts to the diseases that they are likely to encounter in cies still defy classification. A case may be unclas- practice. Opinion was split on the need for a “short sifiable because of an inadequate tissue sample, list,” and a poll taken after the meeting showed a because special studies are not available, because majority of the oncologists favoring one compre- the tissue is poorly preserved, or because even with hensive list with common entities highlighted. complete analysis it does not fit into one of the categories recognized in the classification. For each Summary case, the reason for the inability to classify it should 1. Are clinical groupings of B- and T/NK-cell be stated in the pathology report. Suggested cate- lymphomas necessary or useful? No gories and terminology for unclassifiable cases are 2. Should common entities be indicated in bold? listed in Table 16. Yes 3. Should a short list of common entities be in- CONCLUSIONS cluded for clinicians? No The committee concluded that the approach to the classification of hematologic malignancies pro- Unclassifiable Hematologic Malignancies posed by the International Lymphoma Study Group Even with the advances in immunophenotyping in the REAL classification and adopted now in the and genetic analysis, some hematologic malignan- WHO classification represents a significant advance in our ability to identify and treat specific disease TABLE 13. T-Cell Neoplasms, Disseminated Leukemic entities. This approach leaves room for identifying Types: Variants new entities and subtypes and for incorporating T-cell prolymphocytic leukemia, morphologic variants Small cell new data into diagnostic criteria, disease definition, Cerebriform cell and nomenclature. It has also produced a new and Adult T-cell leukemia/lymphoma (HTLV1ϩ), clinical variants exciting degree of cooperation and communication Acute Lymphomatous between oncologists and pathologists from around Chronic the world that should facilitate accumulation of Smoldering new knowledge and that will hopefully continue in Hodgkin-like the future. After the WHO classification is com-

206 Modern Pathology pleted, it will be important to develop a mechanism lemze, W. Wilson, R.A. Zittoun. Invited Patholo- for updating it, to avoid the confusion that has gists: C.D. Baroni, C. DeWolf-Peeters, N. Hurwitz, J. often resulted from the existence of multiple clas- Jancar, P.M. Kluin, R. Langholm, L. Peterson, D. sifications. Weisenburger, C.L. Willman. The authors acknowledge the generous financial Acknowledgments: The WHO Classification. Steer- support for this meeting provided by Becton- ing Committee: E.S. Jaffe, N.L. Harris, J. Diebold, G. Dickinson, Berlex Laboratories/Schering Berlin, Bris- Flandrin, H.-K. Muller-Hermelink, J. Vardiman. tol Myers Squibb, the Cure for Lymphoma Founda- Emeritus Consultants: C. Berard, K. Lennert. Com- tion, Coulter Corporation, Dako A/S, F. mittee Chairs: R. Brunning, D. Catovsky, A. Feller, Hoffmann-La Roche, Ltd, the Leukemia Clinical Re- T. Grogan, N.L. Harris, D. Knowles, H. Stein, E. search Foundation, the Swiss Federal Office of Pub- Ralfkiaer, J. Vardiman, R. Warnke. Committee lic Health, the National Cancer Institute, the Uni- members: P.M. Banks, R. Bartl, J. Bennett, F. versity of Chicago Cancer Research Center, and the Berger, B. Borisch, J.K.C. Chan, W.C. Chan, G. Del- World Health Organization. sol, W.T. Dura, B. Favara, K.M. Foucar, G. Frizzera, K. Gatter, A. Georgii, D.R. Head, M. Imbert, P.G. Isaacson, M. Kikuchi, R. Kyle, R. Mann, E. Matutes, E. Montserrat, B.N. Nathwani, G. Pallesen, M. REFERENCES Paulli, R.V. Pierre, S. Pileri, M.A. Piris, S. Poppema, 1. Harris NL, Jaffe ES, Stein H, Banks PM, Chan JKC, Cleary M, M. Raphael, J. Said, S.H. Swerdlow, B. Van Camp, et al. A revised European-American classification of lym- L.M. Weiss, D.H. Wright. phoid neoplasms: a proposal from the International Lym- The Clinical Advisory Committee. Chairs: C.D. phoma Study Group. Blood 1994;84:1361–92. Bloomfield, T.A. Lister. Clinical Advisory Commit- 2. World Health Organization classification of neoplastic diseases of the hematopoietic and lymphoid tissues. In: Sobin tee Members: J.O. Armitage, G. Brittinger, A.K. Bur- LH, editor. WHO international histological classification of nett, G.P. Canellos, F. Cavalli, B. Coiffier, J. Con- tumors. Berlin/New York: Springer-Verlag; in press. nors, V.T. DeVita, Jr., V. Diehl, E. Estey, B. Falini, R.I. 3. Bennett J, Catovsky D, Daniel M, Flandrin G, Galton DA, Fisher, C. Gisselbrecht, J.M. Goldman, W. Hidde- Gralnick HR, et al. Proposed revised criteria for the classifi- mann, D. Hoeltzer, R.T. Hoppe, S.J. Horning, J.C. cation of acute myeloid leukemia. Ann Intern Med 1985;103: 620–5. Kluin-Nelemans, A. Levine, D.L. Longo, I. Magrath, 4. Mann R, Berard C. Criteria for the cytologic subclassification S.B. Murphy, R. Ohno, C.-H. Pui, K.R. Rai, S. Rosen- of follicular lymphomas: a proposed alternative method. berg, J.M. Rowe, A.H. Sarris, M.A. Shipp, D.R. Wil- Hematol Oncol 1982;1:187–92.

WHO Classification of Hematological Malignancies (N.L. Harris et al.) 207