Relative Frequencies and Sites of Presentation of Lymphoid Neoplasms in a Community Hospital According to the Revised European-American Classification

Hematopathology / FREQUENCIES AND SITES OF PRESENTATION OF LYMPHOID NEOPLASMS IN A COMMUNITY HOSPITAL

James D. Siebert, MD,1-2 DebraA. Mulvaney, MLT(ASCP) ,J<2 Kari L. Potter, HT(ASCP),1'2 Paul A. S. Fishkin, MD,L3 and Francois J. Geoffroy, MDJ3

Key Words: Lymphoma; Lymphoid neoplasms; REAL classification; Immunophenotype Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021

Abstract In 1994, the International Lymphoma Study Group Relative frequencies for common subtypes in the proposed a revised European-American classification of revised European-American classification of lymphoid lymphoid neoplasms (REAL classification), which catego neoplasms (REAL classification) have been reported. rizes lymphoma variants based on morphologic features, We determined the relative frequencies and sites of immunophenotype, and clinical findings.1 The classifica presentation of REAL subtypes at a 700-bed community tion represents a significant departure from the Working hospital in central Illinois. A database was used to Formulation because of the use of immunophenotype, the identify and prospectively catalogue all newly omission of some Working Formulation entities, and the diagnosed lymphoid neoplasms from July 1, 1995 to recognition of newly characterized lymphoid neoplasms.2 March 1, 1998. The approach to diagnosis and Data about the new classification are being accrued. REAL subtyping incorporated morphologic features, classification subtyping is highly reproducible among immunophenotype, and clinical findings according to expert pathologists, it is most accurate when immunopheno criteria proposed in the REAL classification. Of 347 type is used in conjunction with morphologic features, and lymphoid neoplasms diagnosed, 319 were subtyped in it is clinically relevant for some variants.3-5 Relative the REAL classification. Of these, 261 were B-cell frequencies of REAL classification subtypes also have been neoplasms, 21 were T-cell neoplasms, and 37 were studied. Frequencies of some REAL subtypes have been Hodgkin disease variants. Chronic lymphocytic reported using a modified Working Formulation classifica leukemia/small lymphocytic lymphoma/prolymphocytic tion.6-7 Approximate REAL subtype relative frequencies leukemia, diffuse large cell, and follicle center have been reported by the International Lymphoma Study neoplasms were the most common B-cell subtypes. Group.8 Subtype relative frequencies also have been Large granular lymphocyte leukemia was the most reported by the Non-Hodgkin's Lymphoma Classification common T-cell neoplasm. Nodular sclerosis was the Project (NHLCP), the European Organization for the most common Hodgkin disease variant. The relative Research and Treatment of Cancer (EORTC), and investiga frequencies in a US community hospital setting are tors in Spain using the REAL classification.3'9'10 similar to those reported in other studies. Differences It is unknown whether reported subtype relative are attributable to patient selection criteria, study frequencies are valid for US community hospitals. NHLCP, group geographic location and racial composition, EORTC, and European center data may be influenced by and/or referral patterns. Diverse REAL classification patient selection for participation in clinical trials, overrep- subtypes may be expected in US community hospitals. resentation of T-cell neoplasms in some geographic loca tions and racial groups, and/or referral patterns. It also is unknown whether the REAL classification is practicable in a community hospital.1112 This study was performed to determine the relative frequencies and sites of presentation of REAL classification subtypes and to help determine the

© American Society of Clinical Pathologists Am J Clin Pathol 1999; 111:379-386 379 Siebert et al / FREQUENCIES AND SITES OF PRESENTATION OF LYMPHOID NEOPLASMS IN A COMMUNITY HOSPITAL potential practicability of the REAL classification in a US stains, according to manufacturer's instructions. Heat-induced community hospital. epitope retrieval was not performed for CD43, K, or X stains. For CyclinDl, a 1.0-mmol/L concentration of EDTA solution, pH 8.0, was used in place of commercial buffer for heat- induced epitope retrieval. CyclinDl (1:30) and Tdt (1:10) were Methods titrated in separate validation studies by using commercial OSF Saint Francis Medical Center (SFMC) is a 700-bed diluent (DAKO). Appropriate positive and negative controls community-based, university-affiliated hospital in central were performed in parallel for each assay. Illinois located midway between Chicago, 111, to the north, For 2-color flow cytometry, solid tissue was allocated fresh and Saint Louis, Mo, to the south. A computer database was on sterile normal saline-moistened gauze, manually disaggre used to identify and prospectively catalogue all newly diag gated, washed, and resuspended in phosphate-buffered saline and nosed lymphoid neoplasms at SFMC from July 1, 1995 to 1% sodium azide. Marrow aspirate was collected in EDTA, and Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021 March 1, 1998. The approach to diagnosis included morpho mononuclear cells were obtained by the Ficoll-Hypaque logic evaluation by a hematopathologist (J.D.S.) and on-site (Pharmacia, Piscataway, NJ) density gradient method.14 Cell immunophenotyping with paraffin-section immunohisto- suspension aliquots were mixed with fluorescein isothio- chemistry, flow cytometry, or both. cyanate-conjugated antibody and phycoerythrin-conjugated anti Morphologic evaluation included review of surgical and body, incubated at 4°C with 1% fetal calf serum for 30 minutes, endoscopic tissue biopsy specimens, peripheral blood smears, and fixed with 1 % formaldehyde. Whole blood collected in posterior superior iliac crest bone marrow aspirate and biopsy EDTA, other sterile body fluids, and sterile normal saline suspen specimens procured with a Jamshidi needle, or radiographically sions of cytology samples were processed in a similar fashion, directed cytology smear and tissue core samples procured with and then the RBCs were lysed.15 The monoclonal antibody panel an 18- or 20-gauge needle. Tissue biopsy specimens were used varied slightly depending on case requirements and cell processed routinely in formalin and B5 fixatives, embedded in recovery but most often included CD2/Leu5, CD3/Leu4, paraffin, sectioned at 3 to 4 urn, and stained with H&E. Periph CD4/Leu3a, CD5/Leul, CD8/Leu2a, CD10/CALLA, eral blood and marrow aspirate smears were Wright-Giemsa CDllc/LeuM5, CD14/LeuM3, CD16/Leullc, CD19/Leul2, stained. Marrow biopsy specimens were fixed in B5, decalci CD20/Leul6, CD22/Leul4, CD23/Leu20, CD25/1L2R, fied, and processed similar to tissue biopsy specimens. Cytology CD34/HPCA-2, CD45/Hle-1, CD56/Leul9, u. heavy chain, a smears were air dried or alcohol fixed and Wright-Giemsa or heavy chain, 8 heavy chain, y heavy chain, K light chain, and X Papanicolaou stained, respectively. Cytology needle core light chain (Becton Dickinson, Mountain View, Calif). Anti samples were processed similar to tissue biopsy specimens. body-labeled cell suspensions were analyzed with a flow Paraffin immunostains were performed with tissue cytometer (FACScan, Becton Dickinson) and SimulSET sectioned onto silane-coated (Sigma, St Louis, Mo) slides using (Becton-Dickinson) software. The lymphocyte gate was estab the labeled streptavidin-biotin method.13 Deparaffinized rehy- lished by analysis of forward angle vs right angle light scatter drated sections were placed in 3% hydrogen peroxide for 5 and confirmed by the pattern of CD 14 and CD45 staining. Fluo minutes. After microwave heat-induced epitope retrieval (12 rescein isothiocyanate-labeled and phycoerythrin-labeled minutes on high power) using a commercial antigen retrieval isotype negative control antibodies were used for placement of solution (DAKO, Carpinteria, Calif), sections were rinsed in horizontal and vertical cursors to discriminate positive staining. Tris-buffered saline and incubated in a wet chamber with Cursors were placed such that greater than 98% of lymphocyte manufacturer-diluted primary antibody. Secondary antibody fluorescence for the isotype control was in quadrant 3. and labeling conjugate were applied as directed (DAKO). Additional histochemical studies were performed, when Sections were then placed in diaminobenzidine chromogen for indicated by morphologic features or immunophenotype, 15 minutes, counterstained with hematoxylin, and cover- using air-dried smears and appropriate controls. Stain for slipped. Primary antibodies were selected from a panel acid phosphatase with and without tartrate pretreatment was according to the morphologic differential diagnosis and after performed using standard techniques to substantiate a diag consideration of flow cytometry findings, when applicable. nosis of hairy cell leukemia.16 Indirect immunofluorescence Primary antibodies included CD3 (DAKO), CD15/P12 (Signet for Tdt was performed according to the manufacturer's spec Laboratories, Dedham, Mass), CD20/L26, CD30/BerH2, ifications to help confirm or exclude precursor lymphoblastic CD43/DF-T1, CD45/LCA, CD45RO/UCHL1, K light chain, X leukemia/lymphoma (Supertechs, Bethesda, Md). light chain, bcl-2 protein (DAKO), CyclinDl/2PDllFll Pathologic findings and clinical data were correlated to (Vector Laboratories, Burlingame, Calif), and terminal arrive at a final diagnosis and subtype according to REAL deoxynucleotidyl transferase (Tdt) (DAKO). An enzyme classification criteria. Follicle center neoplasms with a follicular (Proteinase K, DAKO) was used to digest K and X light chain growth pattern that were not interpretable as a manifestation of

380 Am J Clin Pathol 1999; 111:379-386 © American Society of Clinical Pathologists Hematopathology / ORIGINAL ARTICLE mantle cell, marginal zone, or small noncleaved cell (Burkitt criteria in the REAL classification. Twenty-eight lymphoid or high-grade Burkitt-like) lymphoma were graded by using neoplasms could not be subtyped due to noncategorizable blood modified Rappaport criteria.17 For grading, an estimate of the or bone marrow findings (n = 17) or a nontypable cytology percentage of neoplastic follicle area occupied by large cells sample (n = 9) or because findings could not be interpreted as a (centroblasts) was made; if less than 25%, the neoplasm was specific REAL entity (n = 2; 1 a composite lymphoma and 1 a categorized as grade I, if 25% to 50%, grade II, and if more posttransplantation lymphoproliferative disorder). Of 319 than 50%, grade III. T-cell-rich large B-cell lymphomas were subtyped cases, 261 (81.8%) were B-cell neoplasms, 21 (6.6%) diagnosed according to published guidelines and classified in were T-cell neoplasms, and 37 (11.6%) represented Hodgkin the diffuse large B-cell category.18 Peripheral T-cell lymphomas, disease. For each REAL subtype the number of cases, relative unspecified, were subclassified by using the REAL classifica frequency, and sites of presentation are detailed ITable II. tion provisional cytologic grades.1 Chronic lymphocytic leukemia/small lymphocytic Census data for the immediate 3-county area served by lymphoma/prolymphocytic leukemia (CLL/SLL/PLL), diffuse Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021 SFMC and the US were obtained.19 large cell, and follicle center neoplasms were the most common B-cell subtypes. B-CLL/SLL/PLL accounted for 28.2% of subtyped cases and was the most frequent neoplasm mani festing in peripheral blood. The diffuse large B-cell subtype Results accounted for 15.7% of subtyped neoplasms. Most of these We prospectively catalogued 347 newly diagnosed manifested in lymph nodes. Four T-cell-rich large B-cell lymphoid neoplasms. Of these, 319 cases (91.9%) were lymphomas (1 each in lymph node, spleen, gastrointestinal subtyped according to International Lymphoma Study Group tract, and mediastinum) were included in the diffuse large B-cell

ITable II REAL Classification Subtypes: Relative Frequencies and Sites of Presentation in a Community Hospital

Site

Category* Frequency Blood Marrow Lymph Node Spleen Skin GI Tract* Other*

B-cell neoplasms 261 (81.8) 84 (26.3) 42(13.2) 76 (23.8) 11 (3.4) 7 (2.2) 15(4.7) 26 (8.2) CLL/SLL/PLL 90 (28.2) 73 (22.9) 4(1.2) 8 (2.5) 1 (0.3) 1 (0.3) 0 (0.0) 3 (0.9) Lymphoplasmacytoid 3 (0.9) 0 (0.0) 0 (0.0) 3 (0.9) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Mantle cell 17(5.3) 6(1.9) 1 (0.3) 6(1.9) 4(1.2) 0 (0.0) 0 (0.0) 0 (0.0) Follicle center 42(13.2) 2 (0.6) 2 (0.6) 28 (8.8) 0 (0.0) 4(1.2) 1 (0.3) 5(1.6) Marginal zone, nodal (p) and extranodal 11 (3.4) 0 (0.0) 0 (0.0) 2 (0.6) 0 (0.0) 1 (0.3) 7 (2.2) 1 (0.3) Splenic marginal zone (p) 3(0.9) 0 (0.0) 0 (0.0) 0 (0.0) 3(0.9) 0 (0.0) 0 (0.0) 0 (0.0) Hairy cell 4(1.2) 3 (0.9) 1 (0.3) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Plasmacytoma/plasma cell myeloma 22 (6.9) 0 (0.0) 22 (6.9) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Diffuse large cell 50(15.7) 0 (0.0) 8 (2.5) 20 (6.3) 3 (0.9) 0 (0.0) 6(1.9) 13(4.1) Primary mediastinal large cell (p) 5(1.6) 0 (0.0) 0 (0.0) 1 (0.3) 0 (0.0) 0 (0.0) 0 (0.0) 4(1.2) Burkitt 2 (0.6) 0 (0.0) 0 (0.0) 1 (0.3) 0 (0.0) 0 (0.0) 1 (0.3) 0 (0.0) High-grade Burkitt-like (p) 12(3.8) 0 (0.0) 4(1.2) 7 (2.2) 0 (0.0) 1 (0.3) 0 (0.0) 0(0.0) T-cell neoplasms 21 (6.6) 13(4.1) 0 (0.0) 5(1.6) 0 (0.0) 2 (0.6) 1 (0.3) 0 (0.0) Precursor lymphoblastic 1 (0.3) 1 (0.3) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) CLL/PLL 2 (0.6) 2 (0.6) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) T-large granular lymphocyte 7 (2.2) 7 (2.2) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) NK-large granular lymphocyte 3 (0.9) 2 (0.6) 0 (0.0) 1 (0.3) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Mycosis fungoides/Sezary syndrome 2 (0.6) 1 (0.3) 0 (0.0) 0 (0.0) 0 (0.0) 1 (0.3) 0 (0.0) 0 (0.0) Peripheral T-cell, NOS 3 (0.9) 0 (0.0) 0 (0.0) 2 (0.6) 0 (0.0) 0 (0.0) 1 (0.3) 0 (0.0) Anaplastic large cell, T and null cell 1 (0.3) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) 1 (0.3) 0 (0.0) 0 (0.0) Anaplastic large cell, Hodgkin-like (p) 2 (0.6) 0 (0.0) 0 (0.0) 2 (0.6) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Hodgkin disease 37(11.6) 0 (0.0) 0 (0.0) 34(10.7) 0 (0.0) 0 (0.0) 0 (0.0) 3 (0.9) Lymphocyte predominance 4(1.2) 0 (0.0) 0 (0.0) 4(1.2) 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0) Nodular sclerosis 26 (8.2) 0 (0.0) 0 (0.0) 24(7.5) 0 (0.0) 0 (0.0) 0 (0.0) 2 (0.6) Mixed cellularity 6(1.9) 0 (0.0) 0 (0.0) 5(1.6) 0 (0.0) 0 (0.0) 0 (0.0) 1 (0.3) Lymphocyte depleted 1 (0.3) 0 (0.0) 0 (0.0) 1 (0.3) 0 (0.0) 0 (0,0) 0 (0.0) 0 (0.0) Total 319(100.0) 97 (30.4) 42(13.2) 115(36.0) 11 (3.4) 9 (2.8) 16(5.0) 29(9.1) REAL = revised European-American classification of lymphoid neoplasms; GI = gastrointestinal; CLL = chronic lymphocytic leukemia; SLL = small lymphocytic lymphoma; PLL = prolymphocyte leukemia; p = provisional entity or subtype; NK = natural killer cell; NOS = unspecified. * Subtypes for which no case was identified are omitted. Data are given as number (percentage). Totals of percentages for the columns and rows may differ because of rounding. ' Includes Waldeyer ring. + Includes the mediastinum.

© American Society of Clinical Pathologists Am J Clin Pathol 1999;111:379-386 381 Siebert et al / FREQUENCIES AND SITES OF PRESENTATION OF LYMPHOID NEOPLASMS IN A COMMUNITY HOSPITAL category. Age and sex data for these patients were similar to neoplasms. T-cell large granular lymphocyte leukemia was the others in the diffuse large B-cell category. Follicle center most frequent T-cell subtype manifesting in peripheral blood. neoplasms accounted for 13.2% of subtyped cases. Thirty-five Two of 3 peripheral T-cell lymphomas presented in lymph of these 42 neoplasms showed a follicular growth pattern and nodes. One was categorized as a large cell cytologic variant, were graded. Twelve of 35 were categorized as grade 1 (34%), and the other was categorized as a mixed medium and large 15 as grade 2 (43%), and 8 as grade 3 (23%). Seven of 42 cell variant. The third manifested in the gastrointestinal tract follicle center neoplasms were diagnosed with a peripheral (Waldeyer ring) and was categorized as a large cell variant. A blood (n = 2), bone marrow (n = 2), or cytology (n = 3) sample diversity of other REAL T-cell subtypes were identified. and were not graded. All ungraded neoplasms demonstrated Nodular sclerosis was the most common type of small to intermediate-sized cells with condensed chromatin, Hodgkin disease. Most cases of Hodgkin disease manifested variably irregular nuclear contours, and scant to modest in lymph nodes, regardless of subtype. No case of diffuse amounts of cytoplasm. Bone marrow and cytology samples lymphocyte predominance or lymphocyte-rich classic Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021 showed an occasional admixed large transformed cell (centrob- Hodgkin disease was identified. last). Immunophenotypically, all ungraded neoplasms showed The distribution of specimens and immunophenotyping CD5-/CD10+/CD197CD20+/CD22+ B-cells and clear-cut K or methods are detailed for each REAL classification subtype X immunoglobulin light chain restriction. Other REAL B-cell ITable 21. For specimen distribution, most cases were neoplasms were less frequent. Precursor B-lymphoblastic subtyped with a tissue biopsy, peripheral blood, or bone leukemia/lymphoma was the only B-cell subtype not identified. marrow sample. Fourteen were subtyped with cytology smear Large granular lymphocyte leukemia and peripheral T- and needle core samples. Five of these, subtyped in the diffuse cell lymphomas, unspecified, were the most common T-cell large cell category, showed sheets of large transformed B cells

ITable 21 REAL Classification Subtypes: Specimen Distribution and Immunophenotyping Methods

Specimen Distribution Immunophenot yping Method

Category* No. of Specimens Tissue Biopsy Blood/Marrow Cytology FC P-IHC FC and P-IHC

B-cell neoplasms 261 124 126 11 106 77 59 CLL/SLL/PLL 90 11 77 2 80 7 3 Lymphoplasmacytoid 3 3 0 0 3 0 0 Mantle cell 17 10 7 0 9 5 3 Follicle center 42 35 4 3 7 16 15 Marginal zone, nodal (p) and extranodal 11 11 0 0 0 7 4 Splenic marginal zone (p) 3 3 0 0 0 0 3 Hairy cell 4 0 4 0 4 0 0 Plasmacytoma/plasma cell myeloma 22 0 22 0 0 5 2 Diffuse large cell 50 37 8 5 3 27 20 Primary mediastinal large cell (p) 5 5 0 0 0 3 2 Burkitt 2 2 0 0 0 1 1 High-grade Burkitt-like (p) 12 7 4 1 0 6 6 T-cell neoplasms 21 8 13 0 13 3 5 Precursor lymphoblastic 1 0 1 0 1 0 0 CLL/PLL 2 0 2 0 2 0 0 T-large granular lymphocyte 7 o 7 0 7 0 0 NK-large granular lymphocyte 3 1 2 0 2 0 1 Mycosis fungoides/Sezary syndrome 2 1 1 0 1 1 0 Peripheral T-cell, NOS 3 3 0 0 0 1 2 Anaplastic large cell, T and null cell 1 1 0 0 0 0 1 Anaplastic large cell. Hodgkin-like (p) 2 2 0 0 0 1 1 Hodgkin disease 37 34 0 3 0 18 17 Lymphocyte predominance 4 4 0 0 0 1 3 Nodular sclerosis 26 23 0 3 0 14 10 Mixed cellularity 6 6 0 0 0 3 3 Lymphocyte depleted 1 1 0 0 0 0 1 Total 319 166 139 14 119 98 81 REAL = revised European-American classification of lymphoid neoplasms; FC = flow cytometry; P-IHC = paraffin section immunohistochemistry; CLL = chronic lymphocytii leukemia; SLL = small lymphocytic lymphoma; PLL = prolymphocytic leukemia; p = provisional entity or subtype; NK = natural killer cell; NOS = unspecified. * Subtypes for which no case was identified are omitted.

382 Am J Clin Pathol 1999; 111:379-38B © American Society of Clinical Pathologists Hematopathology / ORIGINAL ARTICLE

(centroblasts). Three, subtyped as nodular sclerosis Hodgkin (n = 4) or nodular sclerosis Hodgkin disease (n = 2), were disease, showed CD15+/CD30+/B-cell antigen-negative/T-cell diagnosed in excisional lymph node biopsy specimens that antigen-negative Reed-Sternberg cell variants in a reactive showed typical morphologic attributes of these neoplasms. cellular background with sclerosis. Three were ungraded Patient age and sex data are detailed liable 31. The racial follicle center neoplasms (described in preceding text). Two, distribution of the population in the immediate 3-county area subtyped as B-CLL/SLL, demonstrated typical morphologic served by SFMC is compared with that of the United States features and flow cytometry findings. For immunopheno- •Table 41. This shows that SFMC serves a larger proportion typing methods, 298 of 319 cases were phenotyped with flow of whites than present in the US population as a whole. cytometry, paraffin-section immunohistochemistry, or both methods. Immunophenotype was not determined for 21 subtyped cases. Fifteen of these, subtyped as plasma cell

Discussion Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021 myeloma, were diagnosed in bone marrow samples that showed sheets or abnormal aggregates of plasma cells. The To our knowledge, the present study is the first to remaining 6, subtyped as follicular follicle center lymphoma examine the relative frequency of REAL classification

•Table 31 REAL Classification Subtypes: Age and Sex Distribution

Category* Age Range (y) Mean Age (y)+ M/F

B-cell neoplasms CLL/SLL/PLL 35-87 69 46/44 Lymphoplasmacytoid 51-81 67 0/3 Mantle cell 52-89 72 11/6 Follicle center 37-99 64 22/20 Marginal zone, nodal (p) and extranodal 43-81 67 6/5 Splenic marginal zone (p) 67-80 73 1/2 Hairy cell 60-71 66 3/1 Plasmacytoma/plasma cell myeloma 34-81 67 12/10 Diffuse large cell 30-93 68 27/23 Primary mediastinal large cell (p) 20-81 53 2/3 Burkitt 26-80 — 1/1 High-grade Burkitt-like (p) 20-77 63 7/5 T-cell neoplasms Precursor lymphoblastic 13 1/0 CLL/PLL 48-79 — 2/0 T-large granular lymphocyte 45-87 68 4/3 NK-large granular lymphocyte 67-76 72 2/1 Mycosis fungoides/Sezary syndrome 68-93 — 1/1 Peripheral T-cell, NOS 18-65 41 3/0 Anaplastic large cell, T and null cell 53 — 1/0 Anaplastic large cell, Hodgkin-like (p) 45-51 — 2/0 Hodgkin disease Lymphocyte predominance 11-50 23 3/1 Nodular sclerosis 16-75 36 16/10 Mixed cellularity 27-70 52 6/0 Lymphocyte depleted 75 — 1/0 REAL = revised European-American classification of lymphoid neoplasms; CLL = chronic lymphocytic leukemia; SLL = small lymphocytic lymphoma; PLL = prolymphocytic leukemia; p = provisional entity or subtype; NK = natural killer cell; NOS = unspecified. * Subtypes for which no case was identified are omitted. ' Mean age was calculated for subtypes only if at least 3 cases were identified.

•Table 41 Racial Distributions in Central Illinois and the United States19 Race Central Illinois Population (%)* United States Population (%) White 307,065 (90.5) 199,827,064(73.8) Black 24,990 (7.4) 29,930,524(11.1) Hispanic 3,622(1.1) 21,900,089(8.1) Asian or Pacific Islander 2,711 (0.8) 7,226,986 (2.7) American Indian or Native Alaskan 583 (0.2) 2,015,143(0.7) Other 201 (<0.1) 9,710,156(3.6)

* Data for Peoria, Tazewell, and Woodford counties.

© American Society of Clinical Pathologists Am J Clin Pathol 1999; 111:379-386 383 Siebert et al / FREQUENCIES AND SITES OF PRESENTATION OF LYMPHOID NEOPLASMS IN A COMMUNITY HOSPITAL subtypes in a cohort of newly diagnosed patients in a US neoplasms. First, for B-CLL/SLL/PLL, the higher relative community hospital. All diagnostic samples in all body sites frequency at SFMC (28.2%) may be attributed to different were evaluated so that results would accurately reflect the inclusion criteria, patient age, or both. Cases with the blood diversity of lymphoid neoplasms encountered in routine smear morphology and flow cytometry immunophenotype of practice. The approach to diagnosis incorporated morpho B-CLL/SLL were included in the present study, in accor logic review by an experienced observer (J.D.S.), dance with National Cancer Institute recommendations.20 immunophenotype, and clinical findings. Subclassification Investigators in Spain used similar methods and reported a was performed using International Lymphoma Study Group similar relative frequency (24.9%).10 The NHLCP, in a criteria. By using these methods, 91.9% of all lymphoid multi-institutional study that required tissue biopsy for inclu neoplasms were subtypable in the REAL classification. This sion, reported a lower relative frequency (6.7%).3 The demonstrates the potential practicability of the REAL classi EORTC, in a report of patients selected for participation in fication in a community setting. B-cell neoplasms accounted clinical trials with stage IILTV non-Hodgkin lymphoma diag Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021 for 81.8%, T-cell neoplasms accounted for 6.6%, and nosed in lymph node or Waldeyer ring, reported an even 9 Hodgkin disease accounted for 11.6% of subtyped cases. lower relative frequency (<2%). For patient age, Surveil When methodologic differences are considered, the relative lance, Epidemiology, and End Result (SEER) data have frequencies of REAL classification subtypes in this study are shown that the incidence of B-CLL/SLL increases with 21 similar to those reported by others liable 51. Nevertheless, age. Patients with B-CLL/SLL/PLL at SFMC averaged 69 compared with most other studies, results show higher rela years, slightly higher than the median age reported by 3 9 tive frequencies for B-CLL/SLL/PLL, the high-grade others. - Second, the plasmacytoma/plasma cell myeloma Burkitt-like subtype, and the nodular sclerosis variant among subtype accounted for 6.9% of lymphoid neoplasms, cases of Hodgkin disease. Lower relative frequencies were whereas investigators in Spain reported a relative frequency 10 found for plasmacytoma/plasma cell myeloma and T-cell of 15% for this disorder. The lower relative frequency may neoplasms.3'8"10 be related to the fact that SFMC serves a lower proportion of Several factors could account for the differences in rela blacks, in whom plasma cell myeloma is more common, and a higher proportion of whites, in whom plasma cell myeloma tive frequency that were observed for some B-cell

•Table 51 REAL Classification Subtypes: Relative Frequency Comparisons (%)

Investigator1

Category* SFMC Salar et at10 NHLCP3 Pittaluga et al9 Harris et al8

B-cell neoplasms 82 74 88 96 85 CLL/SLL/PLL 28 25 7 2 5 Diffuse large cell 16 14 31 32 30 Follicle center 13 7 22 46 35 Plasmacytoma/plasma cell myeloma 7 15 NR NR NR Mantle cell 5 2 6 11 5 Marginal zone, nodal (p) and extranodal 3 4 9 5 1-2 High grade Burkitt-like (p) 4 <1 2 NR NR Primary mediastinal large cell (p) 2 <1 2 NR NR Hairy cell 1 <1 NR NR NR T-cell neoplasms 7 9 12 4 15 Large granular lymphocyte 3 1 NR NR NR Peripheral T-cell (NOS, hepatosplenic (p), 2 4 9 4 12 subcutaneous panniculitic (p), angiocentnc, angioimmunoblastic, intestinal, lymphoepithelial, adult T-cell lymphoma/leukemia, plus anaplastic large cell-T/null/Hodgkin-like (p) subtypes) Hodgkin disease 12 10 NR NR NR Nodular sclerosis 8 6 NR NR NR Mixed cellularity 2 2 NR NR NR Lymphocyte predominance 1 <1 NR NR NR REAL = revised European-American classification of lymphoid neoplasms; SFMC = OSF Saint Francis Medical Center; NHLCP = the Non-Hodgkin's Lymphoma Pathologic Classification Project; CLL = chronic lymphocytic leukemia; SLL = small lymphocytic lymphoma; PLL = pro lymphocytic leukemia; p = provisional entity or subtype; NOS = unspecified; NR = not reported. * Comparisons are for REAL subtypes that account for at least I % of SFMC lymphoid neoplasms. SFMC data and data from reference 10 are for all sites. Data from reference 3 are for non-Hodgkin lymphomas diagnosed by tissue biopsy. Data from reference 9 are for stage ITI/IV non-Hodgkin lymphomas diagnosed in lymph node or Waldeyer ring.Dat a from reference 8 are approximations for nodal presentations of non-Hodgkin lymphomas in US and European adults.

384 Am J Clin Pathol 1999; 111:379-386 © American Society of Clinical Pathologists Hematopathology / ORIGINAL ARTICLE is less common, than present in the United States.1922 The proportions of these neoplasms. Nodular sclerosis Hodgkin racial distribution of patients in Spain was not reported.10 In disease is relatively more common among whites.31 This addition to race, geographic location may influence the inci racial group accounts for 90.5% of the population in the dence of plasma cell myeloma.23 Third, the high-grade vicinity of SFMC, whereas whites account for 73.8% of the Burkitt-like provisional subtype accounted for 3.8% of US population.19 Mixed cellularity Hodgkin disease is more lymphoid neoplasms, whereas others have reported relative common in Hispanics.32-33 This racial group accounts for frequencies that range only up to 2%.3-8-9 In the present 1.1% of the population in the vicinity of SFMC but 8.1% of study, the high-grade Burkitt-like subtype was reserved for the US population.19 highly proliferative B-cell neoplasms that demonstrated From a practical perspective, the recognition of many morphologic features intermediate between large cell and REAL classification subtypes was greatly aided by Burkitt lymphoma. None of these cases fulfilled criteria for immunophenotyping. In addition to facilitating identification of lymphoblastic lymphoma or blastic mantle cell the major B-cell, T-cell, and Hodgkin lymphoma groupings, Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021 1 24 lymphoma. - An increased frequency of AIDS in these these techniques were crucial in characterizing mantle cell patients, which may have explained an increased relative lymphoma, T-cell lymphoma subtypes, and large B-cell 25 26 frequency, was not found (data not shown). - Because the lymphomas that contained a high proportion of reactive T-cells. high-grade Burkitt-like subtype is reproducible among The present study attempted to further define the rela experts in only 53% of cases after consideration of morpho tive frequencies and sites of presentation of REAL classifica logic, immunologic, and clinical variables, some difference tion neoplasms. Results show that a diversity of REAL 3 in relative frequencies may be expected. subtypes may be expected to be encountered in US commu Results also show that REAL T-cell neoplasms nity hospitals. Results also demonstrate the potential practi accounted for 6.6% of subtyped lymphoid neoplasms. Most cability of the REAL classification in a community hospital others have documented relative frequencies that range from setting. The feasibility of the new classification may be 9% to 15%.3-8-9 This difference may be related to patient further defined by accuracy and reproducibility studies that selection, referral patterns, or the geographic location and compare REAL classification diagnoses of community and racial distribution of participating centers. For patient selec university-based hospitals, as well as those that compare tion, some have excluded peripheral blood samples and eval diagnoses of different community hospitals. uated only non-Hodgkin lymphomas, whereas others have considered only non-Hodgkin lymphomas in lymph node or From the 'OSF Saint Francis Medical Center and the University Waldeyer ring.3-9 When these exclusions are applied to of Illinois College of Medicine, Departments of2Pathology and 3 results of the present study, 8 (4.3%) of 185 and 6 (7%) of 82 Laboratory Medicine and Hematology and Oncology, Peoria, cases are T-cell neoplasms, respectively. Because these Illinois. exclusions do not substantially increase the frequency of T- Address reprint requests to Dr Siebert: Saint Francis cell neoplasms, referral patterns and/or the geographic loca Medical Center, Department of Pathology and Laboratory Medicine, 530 NE Glen Oak Ave, Peoria, 1L 61637. tion and racial distribution of participating centers may better explain the discrepant relative frequency. Referral patterns could have a role because uncommon lymphoma variants may have been preferentially referred to university-based References health care settings. We did not study the potential effect of 1. Harris NL, Jaffe ES, Stein H, et al. A revised European- referral patterns. The geographic location and racial distribu American classification of lymphoid neoplasms: a proposal tion of participating centers also could have a role, because from the International Lymphoma Study Group. Blood. some REAL classification T-cell subtypes are more common 1994;84:1361-1392. in Asia and the middle east.27*30 In the immediate area 2. The Non-Hodgkin's Lymphoma Pathologic Classification Project. National Cancer Institute sponsored study of served by SFMC, Asians account for less than 1% of the classifications of non-Hodgkin's lymphomas. Cancer. population.19 Despite the lower relative frequency of T-cell 1982;49:2112-2135. neoplasms, most REAL T-cell subtypes were encountered. 3. The Non-Hodgkin's Lymphoma Classification Project. A clinical evaluation of the International Lymphoma Study For Hodgkin disease, the nodular sclerosis type Group classification of non-Hodgkin's lymphoma. Blood. accounted for 26 (70%) of 37 cases, whereas the mixed 1997;89:3909-3918. cellularity type accounted for 6 (16%) of 37 cases. In 4. Fisher RI, Dahlberg S, Nathwani BN, et al. The clinical contrast, SEER Hodgkin disease data, representative of the analysis of two indolent lymphoma entities: mantle cell United States as a whole, categorized 58% of cases as lymphoma and marginal zone lymphoma (including the 31 mucosal associated lymphoid tissue and monocytoid B-cell nodular sclerosis and 23% as mixed cellularity. The racial subcategories): a Southwest Oncology Group study. Blood. composition of central Illinois may explain the difference in 1995;85:1075-1082.

5. Melnyk A, Rodriguez A, Pugh WC, et al. Evaluation of the 19. United States Government Census Bureau, Chicago, 111, revised European-American lymphoma classification confirms 1990. the clinical relevance of immunophenotype in 560 cases of 20. Cheson BD, Bennett JM, Grever M, et al. National Cancer aggressive non-Hodgkin's lymphoma. Blood. Institute-sponsored Working Group guidelines for chronic 1997;89:4514-4520. lymphocytic leukemia: revised guidelines for diagnosis and 6. Carbone A, Francheschi S, Gloghini A, et al. Pathologic and treatment. Blood. 1996;87:4990-4997. immunophenotypic features of adult non-Hodgkin's 21. Greiner TC, Medeiros L], Jaffe ES. Non-Hodgkin's lymphomas by age group. Hum Pathol. 1997;28:580-587. lymphoma. Cancer. 1995;75:370-380. 7. Glass AG, Karnell LH, Menck HR. The National Cancer 22. Brown LM, Pottern LM, Silverman DT, et al. Multiple Data Base report on non-Hodgkin's lymphoma. Cancer. myeloma among blacks and whites in the United States: role 1997;80:2311-2320. of cigarettes and alcoholic beverages. Cancer Causes Control. 8. Harris NL. Principles of the revised European-American 1997;8:610-614. lymphoma classification (from the International Lymphoma 23. Franceschi S, Levi F, La Vecchia C, et al. Comparison of Study Group). Ann Oncol. 1997;8(suppl 2):S11-S16. cancer mortality trends in major European areas, 1960-1989. 9. Pittaluga S, Bijnens L, Teodorovic I, et al. Clinical analysis of Eur) Cancer Prev. 1994;3:145-206. Downloaded from https://academic.oup.com/ajcp/article/111/3/379/1758571 by guest on 27 September 2021 670 cases in two trials of the European Organization for the 24. Lardelli P, Bookman MA, Sundeen J, et al. Lymphocytic Research and Treatment of Cancer lymphoma cooperative lymphoma of intermediate differentiation: morphologic and group subtyped according to the revised European-American immunophenotypic spectrum and clinical correlation. Am ] classification of lymphoid neoplasms: a comparison with the SurgPathol. 1990;14:753-763. Working Formulation. Biood. 1996;87:4358-4367. 25. Raphael MM, Audouin], Lamine M, et al. 10. Salar A, Fernandez de Sevilla A, Romagosa V, et al. Immunophenotypic and genotypic analysis of acquired Distribution and incidence rates of lymphoid neoplasms immunodeficiency syndrome-related non-Hodgkin's according to the REAL classification in a single institution: a lymphomas: correlation with histologic features in 36 cases. prospective study of 940 cases. Eur J Haematol. AmJCiinPatnoL 1994;101:773-782. 1997;59:231-237. 26. Levine A. Lymphoma in acquired immunodeficiency 11. Rosenberg SA. Classification of lymphoid neoplasms. Biood. syndrome. Semin Oncol. 1990;17:104-112. 1994;84:1359-1360. 27. Kadin ME, Berard CW, Nanba K, et al. Lymphoproliferative 12. Al-Saleem T Classification of lymphoid neoplasms between diseases in Japan and western countries. Hum Pathol. the hematopathologists and the "common person." Blood. 1983;14:745-772. 1995;85:1971-1980. 28. Al-Mondhiry H. Primary lymphomas of the small intestine: 13. Guesdon JL, Ternynck T, Avrameas S, et al. The use of avidin- East-West contrast. Am J Hematoi. 1986;22:89-105. biotin interaction in immunoenzymatic techniques. J Histo- chemCytochem. 1979;27:1131-1139. 29. Domizio P, Owen RA, Shepherd NA, et al. Primary lymphoma of the small intestine: a clinicopathologic study of 14. Knowles DM. Organization and operation of a 119 cases. Am] Surg Pathol. 1993;17:429-442. hematopathology laboratory. In: Knowles DM, ed. Neoplastic Hematopathology. Baltimore, Md: Williams 6k Wilkins; 30. Matutes E, Catovsky D. Mature T-cell leukemias and 1992:358 (Appendix 1). leukemia/lymphoma syndromes: review of our experience in 175 cases. Leuk Lymphoma. 1991;4:81-91. 15. Knowles DM. Organization and operation of a hematopathology laboratory. In: Knowles DM, ed. Neoplastic 31. Medeiros LJ, Greiner TC. Hodgkin's disease. Cancer. Hematopathology. Baltimore, Md: Williams & Wilkins; 1995;75:357-369. 1992:358 (Appendix 2). 32. Hu S, Hufford S, Lukes R, et al. Third-world Hodgkin's 16. Yam LT, Li CY, Lam KW Tartrate-resistant acid phosphatase disease at Los Angeles County-University of Southern isoenzyme in the reticulum cells of leukemic reticulo- California Medical Center. J Clin Oncol. 1988;6:1285-1292. endotheliosis. N Engl] Med. 1971;284:357-360. 33. Chang KL, Albujar PF, Chen YY, et al. High prevalence of 17. Rappaport H. Tumors of the Hematopoietic System. Atlas of Epstein-Barr virus in the Reed-Sternberg cells of Hodgkin's Tumor Pathology, Fascicle 8, Section 3. Washington, DC: disease occurring in Peru. Biood. 1993;81:496-501. Armed Forces Institute of Pathology; 1966:97-161. 18. Chan WC. T-cell rich B-cell lymphoma: what is new? what is cool? Am ] Clin Pathol. 1997;108:489-490.