Relative Frequencies and Sites of Presentation of Lymphoid Neoplasms in a Community Hospital According to the Revised European-American Classification
Total Page:16
File Type:pdf, Size:1020Kb
Hematopathology / FREQUENCIES AND SITES OF PRESENTATION OF LYMPHOID NEOPLASMS IN A COMMUNITY HOSPITAL Relative Frequencies and Sites of Presentation of Lymphoid Neoplasms in a Community Hospital According to the Revised European-American Classification 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