Hematopathology / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS

Transcription

1 Hematopathology / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS Histopathologic Features of Splenic Small B-Cell Lymphomas A Study of 42 Cases With a Definitive Diagnosis by the World Health Organization Classification Rina Kansal, MD, Charles W. Ross, MD, Timothy P. Singleton, MD, William G. Finn, MD, and Bertram Schnitzer, MD Key Words: Splenic lymphoma; Marginal zone; Small B-cell; Mantle cell lymphoma; Follicular lymphoma; Chronic lymphocytic leukemia; Small lymphocytic lymphoma; WHO classification DOI: /HWG084N3F3LRJ8XB Abstract We studied 42 cases of splenic small B-cell lymphoma (SBL) (21 women, 21 men; aged years; median, 65 years) with a definitive diagnosis by the World Health Organization classification: chronic lymphocytic leukemia (CLL), 8; mantle cell lymphoma (MCL), 9; follicular lymphoma (FL), 12; marginal zone lymphoma, 13 (splenic [SMZL], 12; extranodal [EMZL], 1). Splenectomy was performed for diagnosis or therapy; splenic weights were 0.2 to 3.8 kg (median, 1.4 kg). In general, splenic SBLs showed white pulp (WP) expansion; morphologic features of the nodules recapitulated the corresponding lymph node histopathologic features. Marginal zones were observed commonly in SMZL and FL, may be present in MCL involving the spleen, and may be seen in hilar lymph nodes (HLNs) in SBLs other than SMZL. FL may simulate SMZL and can be distinguished by the presence of neoplastic follicles and HLN morphologic features. Extracellular hyaline deposits (EH) are common in FL and SMZL. MCL typically shows WP expansion by a monotonous small lymphocytic infiltrate, without diffuse red pulp (RP) infiltration or EH; leukemic MCL may show RP infiltration. Splenic morphologic features in CLL vary in WP or RP dominance; marginal zones usually are not observed in CLL. Malignant lymphomas commonly involve the spleen secondarily as part of generalized disease. 1,2 Occasionally, lymphomas manifest with prominent splenomegaly, with or without overt evidence of lymphadenopathy, and a splenectomy is performed for diagnostic purposes. 3-7 Small B-lymphocytic neoplasms are conceptualized as clonal proliferations of mature small B-lymphocytes, with different subtypes arising from various stages in the life of a mature B cell. 8 The differential diagnosis of small B-cell lymphomas (SBLs) involving the spleen can be difficult, especially in cases with prominent splenomegaly only, in which no additional diagnostic interpretation can be made from a lymph node biopsy specimen. The significant differences in therapy and prognosis within various subclasses of SBLs mandate precise diagnostic subclassification. 9 While a number of studies have focused on marginal zone lymphomas of the spleen, few studies have compared the different subtypes of splenic SBL after the description of splenic marginal zone lymphomas In the most recent study by Piris et al, 12 the primary diagnosis in 11 of 33 study cases was made by peripheral blood examination alone, with no supportive flow cytometric immunophenotypic or lymph node histopathologic evidence for the diagnoses of 9 chronic lymphocytic leukemias and 2 mantle cell lymphomas. The objective of our study was to define the morphologic features of SBL involving the spleen by studying the cases of splenic SBL in which a definitive diagnosis and subclassification was established according to the recent World Health Organization (WHO) classification, after integrating all available clinical and pathologic information. Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB 335

2 Kansal et al / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS Materials and Methods Case Selection The surgical pathology laboratory information system at the University of Michigan, Ann Arbor, was searched for lymphomas involving the spleen from January 1980 to May Sections of all cases of splenic lymphomas retrieved in the search were reviewed (R.K, B.S.) to exclude cases of Hodgkin lymphoma, diffuse large B-cell lymphoma, T-cell neoplasms, and follicular lymphoma, grade 3, involving the spleen. The case cohort reflects sequential patients with splenic SBL who received treatment at the University of Michigan. The diagnostic paraffin-embedded, B5- or formalin-fixed, H&E-stained sections of the spleen, lymph node or bone marrow biopsies, immunohistochemical stains, peripheral blood smears, and corresponding flow cytometric data and clinical information were reviewed. The diagnostic histologic sections in all cases were reviewed by at least 3 pathologists (R.K., C.W.R., B.S.). The dot plots for all flow cytometric immunophenotypic analyses were reviewed by at least 2 pathologists (R.K., C.W.R.). All cases included in the study were classified according to currently accepted criteria, 13,14 with specific inclusion criteria as follows: B-Cell Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma All patients with B-cell chronic lymphocytic leukemia (CLL) had absolute lymphocytosis and involvement of peripheral blood and bone marrow at the time of diagnosis, with morphologic and flow cytometric immunophenotypic features characteristic of CLL. By flow cytometry, features typical for CLL/small lymphocytic lymphoma (SLL) included expression of CD5, dim or absent CD11c, usually dim CD20, dim or absent CD22, CD23, dim surface immunoglobulin expression, and dim or absent FMC7 on the neoplastic cells. 15 One case of CLL with increased (18%) prolymphocytes 14,16,17 was included in this category. Cases classified as SLL had a diagnostic lymph node biopsy showing histologic features on H&E-stained sections diagnostic of this subclassification. Mantle Cell Lymphoma All cases had a peripheral or hilar lymph node (HLN) biopsy specimen available for review, with features diagnostic of mantle cell lymphoma (MCL). The lymphoma cells were cyclin D1+ by immunohistochemical analysis in all cases. Follicular Lymphoma All cases had a peripheral or HLN biopsy specimen available for review, with histopathologic features diagnostic of follicular lymphoma (FL), grades 1 or 2. Marginal Zone Lymphoma Patients given a diagnosis of splenic marginal zone lymphoma (SMZL) underwent diagnostic splenectomy for extramedullary tissue diagnosis and subclassification and had no involvement of extranodal sites, as determined by review of all available information in the patients medical records. One MZL case with previous involvement of extranodal sites was classified as extranodal MZL (EMZL) subsequently involving the spleen. We did not find any case of splenectomy performed for a nodal MZL in our study. Splenic histopathologic features on H&E-stained sections were as described in the WHO classification. 14 In addition, all cases had peripheral blood or HLN specimens available for review, with features consistent with or typical of MZL. In all selected cases, features specific for other subtypes of SBL were absent, including neoplastic follicles or proliferation centers in histologic sections, circulating cleaved lymphoma cells evident by peripheral blood smear morphologic examination, and bone marrow morphologic and clinical features typical for lymphoplasmacytic lymphoma/ Waldenström macroglobulinemia. The neoplastic cells in all cases were negative for CD5 and for CD10 by immunoperoxidase stains on paraffin-embedded sections or by multicolor flow cytometric immunophenotypic analysis. CD23 positivity in cases with features otherwise consistent with MZL and with absence of features that would suggest CLL/SLL did not exclude the diagnosis of MZL; 2 such cases with CD23 positivity by flow cytometry were included. In all cases with available sections, the lymphoma cells were negative for cyclin D1 by immunohistochemical analysis; 3 cases without available sections for a cyclin D1 stain were included. We did not find any cyclin D1+ case in our study with histologic features typical for MZL. Excluded Cases Only 1 patient with a definitive diagnosis of lymphoplasmacytic lymphoma/waldenström macroglobulinemia underwent splenectomy in our series of cases, and this diagnostic category was, therefore, excluded. Cases with unavailable or inadequate diagnostic material for review were excluded; these cases included 3 suspected FLs and 1 suspected lymphoplasmacytic lymphoma. One SBL case involving peripheral blood and bone marrow, with clinical and pathologic features suggesting SMZL, was excluded because the resected spleen consisted predominantly of nonviable tissue. Pathologic Features and Immunohistochemical Analysis The weight of the spleen and significant macroscopic findings were recorded from the original pathology reports. Histopathologic features studied in the spleen included the presence and extent of white pulp (WP) expansion, presence 336 Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB

3 Hematopathology / ORIGINAL ARTICLE or absence of residual germinal centers (GCs), extracellular hyaline deposits in the WP, presence of a marginal zone (MZ) pattern, presence of red pulp (RP) infiltration (nodular or diffuse), and the cytologic composition of the neoplastic cell population. Splenic HLN sections were reviewed, when available, to assess nodal architecture, preservation of sinuses, presence of a MZ pattern, and the cytologic features of neoplastic cells. In cases with involvement of the peripheral blood or bone marrow, the available smears and sections were reviewed. The H&E- and immunohistochemically stained sections of all diagnostic lymph node biopsies were reviewed to confirm the diagnoses. Additional paraffin section immunoperoxidase stains were performed, if necessary, to confirm the diagnoses. A cyclin D1 immunoperoxidase stain was performed in all cases of MCL and MZL with available sections. All immunostains used an automated stainer (Ventana, Tucson, AZ). The antibodies (clone, source) and antigen retrieval methods used are listed in Table 1. For antigen retrieval, sections were microwaved in a 10-mmol/L concentration of citrate buffer, ph 6.0, for 10 or 15 minutes or microwaved in a 0.25-mol/L tris(hydroxymethyl)aminomethane 0.1- mmol/l EDTA solution, ph 9.0, for 15 minutes, or digested with protease (Protease 1, Ventana) for 16 minutes. Flow Cytometric Immunophenotyping Flow cytometric immunophenotypic analysis of lymphoma cells was performed in 44 specimens from 32 patients in the study. For 9 patients (CLL, 1; MCL, 3; FL, 1; SMZL, 4) more than 1 specimen source was analyzed by flow cytometry. The 44 specimens analyzed included 16 peripheral blood samples (CLL, 5; SMZL, 5; MCL, 4; FL, 2), 14 spleens (SMZL, 8; CLL, 2; MCL, 2; FL, 2), 11 bone marrow aspirates (SMZL, 4; MCL, 3; CLL/SLL, 2; FL, 1; EMZL, 1), 2 lymph nodes (FL, 1; MCL, 1), and 1 pleural fluid sample (FL). The specimens were analyzed using an EPICS XL-MCL flow cytometer (Beckman Coulter, Miami, FL), using multicolor analysis and previously described methods 15 for the majority of cases in the study. Cells from lymph nodes and splenic tissue were obtained by mechanical disaggregation. The antibodies with the directly conjugated fluorochromes used for flow cytometry included the following: CD5/phycoerythrin (PE; Becton Dickinson, San Jose, CA), CD10/fluorescein isothiocyanate (FITC; Becton Dickinson), CD11c/PE (Immunotech, Miami, FL), CD19/PE (Immunotech), CD20/FITC (Becton Dickinson), CD22/FITC (Immunotech), CD23/PE (Becton Dickinson), CD25/FITC (Immunotech), CD45/ECD (Immunotech), CD103/PE (Immunotech), FMC7/FITC (Immunotech), κ/fitc (Becton Dickinson), and λ/pe (Becton Dickinson). A surface antigen was considered positive if at least 20% of the neoplastic cells showed flow cytometric immunofluorescence in excess of an isotype-matched negative control. The flow cytometry dot plots were reviewed in all cases. All cases had immunoglobulin light chain restriction diagnostic of lymphoma. For CD20 and surface immunoglobulins, the intensity of fluorescence was reported as dim if the mean fluorescence intensity was within the first positive logarithmic decade and as bright if the mean fluorescence intensity was in the second or third logarithmic decade. Statistical Analysis The Fisher exact probability test was applied to test for significant histologic differences in the SBL subtypes, using SPSS for Windows, version 10.1 (SPSS, Chicago, IL). The subtypes analyzed were MZL vs CLL/SLL, MZL vs MCL, and MZL vs FL. For the purpose of statistical analysis, the EMZL case was combined with the 12 SMZL cases. The histologic variables were considered positive or negative; focal positivity, as in focally effaced HLN sinuses, and rarely observed histologic variables were considered positive. A P value of less than.05 indicated a statistically significant difference. Clinical Information Clinical information was obtained by medical chart review in all cases. Patient age and sex, stage of disease, Table 1 Immunoperoxidase Stains Antigen Antibody Clone Source Pretreatment CD3ε Polyclonal DAKO, Carpinteria, CA 15 min, citrate, microwave CD5 4C7 Novocastra, Burlingame, CA Tris-EDTA, microwave CD10 56C6 Novocastra Tris-EDTA, microwave CD20 L26 DAKO 15 min, citrate, microwave CD23 BU38 The Binding Site, San Diego, CA 16 min, Protease 1 (Ventana, Tucson, AZ) CD43 L60 Becton Dickinson, San Jose, CA 15 min, citrate, microwave Cyclin D1 AM29 Zymed, South San Francisco, CA Tris-EDTA, microwave κ Polyclonal DAKO 10 min, citrate, microwave λ Polyclonal DAKO 10 min, citrate, microwave Tris, tris(hydroxymethyl)aminomethane. Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB 337

4 Kansal et al / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS presence of involvement of lymph nodes and bone marrow, and the clinical indications for splenectomy were recorded in all cases, with the available follow-up information. Diagnostic intent included splenectomies performed for specific subclassification of lymphoma, following a suspected or confirmed diagnosis of a lymphoproliferative disorder by peripheral blood or bone marrow examination. Results Pathologic Features Forty-two cases that met the aforementioned criteria were included. Table 2 shows a summary of the pathologic features, including the median spleen weights, numbers of cases with the studied histologic features in the spleen and splenic HLNs, and numbers of cases with a diagnostic lymph node biopsy or supportive flow cytometric immunophenotypic analysis, in the different lymphoma subtypes. The splenic gross features were available for 31 cases. For all cases, the spleen weight range was 0.2 to 3.8 kg (median, 1.4 kg). For each subtype of lymphoma, the spleen weight ranges were as follows: CLL, 0.2 to 1.4 kg; MCL, 0.9 to 3.8 kg; FL, 0.6 to 1.9 kg; and MZL, 0.6 to 3.1 kg. By gross examination, prominence of the WP was reported in 28 of 31 cases (CLL, 3/6; MCL, 7/7; FL, 8/8; MZL, 10/10). A miliary pattern was reported in 23 cases (CLL, 3; MCL, 7; FL, 6; MZL, 7). Occasional WP nodules up to 0.6 cm in diameter were noted in 1 case of CLL. Multiple fleshy nodules up to 2.5 cm in diameter were described in 2 cases of MCL. WP nodules in MZL were reported to measure up to 0.5 cm in diameter. Prominent expansion of the RP was reported in 3 cases of CLL. Histopathologic Features B-Cell Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma (n = 8) Microscopically, there was expansion of the WP Image 1A in all cases, primarily by small, mature-appearing lymphocytes admixed with scattered, larger cells with round, vesicular nuclei and distinct, central nucleoli (paraimmunoblasts) Image 1B. In cases 1 and 5, these larger cells were clustered into aggregates referred to as proliferation centers, often highlighted by an immunohistochemical stain for CD23. In case 4, the lymphoid infiltrate included a few lymphocytes with angulated nuclei in addition to small lymphocytes with round nuclear contours. Infiltration of the RP was present in all cases Image 1C and Image 1D and was predominantly nodular in 1 case, predominantly diffuse in 3 cases, and nodular and diffuse in 4 cases. A rare GC was identified in 1 case, with occasional GCs in 2 additional cases. Extracellular hyaline deposition usually was absent. MZs were not observed in the spleen or splenic HLNs. All available splenic HLNs showed at least partial effacement of the nodal architecture, with complete effacement in 4 cases, due to an infiltrate composed primarily of small lymphocytes admixed with scattered paraimmunoblasts. Proliferation centers were identified in HLNs from 2 cases, one of which also had the same feature in the spleen. Table 2 Summary of Pathologic Features in Splenic Small B-Cell Lymphomas Histopathologic Features Present in Spleen Red Pulp Infiltration Splenic Hilar Lymph Nodes Median Spleen Sinuses LN Diagnosis Weight (kg) WP GCs EH MZ Nodular Diffuse Positive * Effaced * MZ * FC Diagnosis CLL/SLL (n = 8) /7 7/7 0/7 8 3 MCL (n = 9) # 6/6 4/6, focal 0/6 6 7 FL (n = 12) /10 10/10 4/ SMZL (n = 12) /10 2/10 8/ EMZL (n = 1) /1 0/1 1/1 1 0 Total (n = 42) /34 23/34 13/ CLL/SLL, B-cell chronic lymphocytic leukemia/small lymphocytic lymphoma; EH, extracellular hyaline deposits; EMZL, extranodal marginal zone lymphoma; FC, cases with flow cytometric immunophenotypic analysis; FL, follicular lymphoma, grade 1 or 2; GCs, germinal centers; HLN, hilar lymph node; LN diagnosis, cases with a lymph node biopsy diagnostic of specific subtype of small B-cell lymphoma; MCL, mantle cell lymphoma; MZ, marginal zone pattern; SMZL, splenic marginal zone lymphoma; WP, white pulp expansion. * Data are given as positive/available HLNs; cases with effaced sinuses in HLNs/positive HLNs; and cases with MZ in HLNs/positive HLNs, respectively. Prominent WP in all cases, with massive expansion in 6 cases. Minute, in rare nodules in 2 cases, and in multiple nodules in 1 case. In rare nodules. Thin zones, in occasional nodules. Often prominent. # Typically minimal; noticeable in 2 cases with leukemic involvement. 338 Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB

5 Hematopathology / ORIGINAL ARTICLE A B C D Image 1 B-cell chronic lymphocytic leukemia (CLL). A, Spleen with irregular white pulp expansion, absence of marginal zones (H&E, 25). B, Splenic white pulp with scattered paraimmunoblasts (H&E, 400). C, Red pulp infiltration, nodular (arrow), by small lymphocytes (H&E, 400). D (Case 6), CLL with increased prolymphocytes. Spleen, red pulp infiltration, diffuse, by small lymphocytes and paraimmunoblasts (H&E, 640); inset, circulating prolymphocytes in the same case (Wright-Giemsa, 1,000) Mantle Cell Lymphoma (n = 9) Microscopically, the WP was expanded in all cases, with massive expansion in 6 cases Image 2A in the form of variably sized, focally coalescent nodules. Near-total effacement of architecture by these massive nodules with minimal intervening RP corresponded to the macroscopic fleshy nodules reported in 2 cases. Infiltration of the RP by lymphoma was present in the form of small nodules Image 2B, typically with minimal diffuse infiltration. In cases 16 and 17, diffuse RP infiltration was noted; both of these cases had a moderate absolute peripheral blood lymphocytosis as well. GCs were very small (Image 2A) and rare in 2 cases and were present in multiple nodules in 1 case. In this latter case, a triphasic pattern was present with a central GC, an intermediate darker zone of small lymphocytes, and an outer lighter MZ. Extracellular hyaline was noted within a GC in 1 case and within a neoplastic nodule in 1 case. Thin MZs were present in occasional nodules in the spleen in 3 cases. The neoplastic cells in all cases included predominantly small lymphocytes with scant cytoplasm and variably irregular nuclear contours, relatively coarsely clumped chromatin, and indistinct nucleoli characteristic of MCL Image 2C. Scattered or clustered pale histiocytes were present within the neoplastic infiltrate in all cases. One case classified as the blastoid variant of MCL had larger cells with more finely distributed nuclear chromatin and prominent mitotic activity. Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB 339

6 Kansal et al / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS A B C D Image 2 Mantle cell lymphoma. A, Spleen with massive white pulp expansion, residual germinal center (arrow), absence of extracellular hyaline, and monotonous population of small lymphocytes within the nodule (H&E, 50). B, Red pulp infiltration, nodular (arrow) (H&E, 640). C, White pulp, cytologic features characteristic of mantle cell lymphoma (H&E, 1,000). D, Hilar lymph node with mantle zone pattern of involvement by lymphoma (H&E, 50). All available splenic HLNs contained a cellular infiltrate consistent with MCL, and all showed at least partial preservation of sinuses. A mantle zone pattern was observed occasionally in the HLNs, with a residual GC surrounded by the neoplastic cells Image 2D. A MZ was not observed in the HLNs, even in 2 cases in which a MZ had been observed in the spleen sections. Follicular Lymphoma (n = 12) Microscopically, all cases showed prominence of the WP, occasionally with variably shaped, irregular, coalescent nodules. Neoplastic follicle formation was identified in all cases. Rare GCs were identified in 1 case. Extracellular hyaline was present within neoplastic nodules in 9 (75%) of 12 cases Image 3A and Image 3B. MZs were observed in half of the cases and often were prominent (Image 3A). The cytologic composition of the MZs was variable from case to case, with a variable number of large cells admixed with small lymphocytes, including cleaved cells. The majority of cases (10/12 [83%]) had small RP nodules composed of cells similar to those in the neoplastic follicles. The RP involvement was mainly diffuse in 2 cases, with 3 additional cases showing both nodular and diffuse RP infiltration. All available HLNs showed neoplastic follicles characteristic of FL with effacement of nodal architecture of sinuses Image 3C. Five of 6 cases with MZs in the spleen 340 Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB

7 Hematopathology / ORIGINAL ARTICLE A B C D Image 3 Follicular lymphoma. A, Spleen with uniform white pulp expansion, prominent marginal zones, and frequent extracellular hyaline deposits (H&E, 25). B, White pulp with extracellular hyaline deposition between small cleaved cells, and surrounding marginal zone at the corner of the photomicrograph (arrows) (H&E, 400). C, Hilar lymph node with effaced sinuses (H&E, 25). D, Hilar lymph node with a rim of marginal zone (outlined by arrows) surrounding the neoplastic follicle (H&E, 400). had HLNs available for review, and 4 of these HLNs also showed MZs Image 3D. Marginal Zone Lymphoma (n = 13) Twelve of 13 cases were classified as SMZL and 1 case (case 35) as EMZL involving the orbit and breast, with subsequent involvement of the spleen 14 years after the initial diagnosis. One case (case 31) had a cervical lymph node involved by MZL 10 years after the splenectomy. Clinically, 25% of patients with SMZL had a history of an immunologic disorder (cases 30, 36, and 42). Microscopically, all cases showed WP expansion Image 4A, usually in the form of round nodules and often with numerous smaller nodules in the RP. Very uniform-sized nodules were noted in case 32. MZs were observed in the spleen in 11 of 13 cases. A biphasic pattern was identified in at least a few nodules in the majority of cases, ie, a central variably thick zone of small lymphocytes surrounded by a variably thick paler zone (MZ) of small lymphocytes with more abundant cytoplasm Image 4B. Small residual GCs within these nodules, present in 11 of 13 cases, occasionally imparted a triphasic appearance. The proportion of small and large lymphoid cells within the MZs was variable in different cases. One case with limited sections for review (case 30) showed predominantly small nodules in the RP and few enlarged WP nodules with MZs. Extracellular hyaline Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB 341

8 Kansal et al / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS A B C D Image 4 Splenic marginal zone lymphoma. A, Spleen with expanded white pulp nodule showing a residual germinal center (arrow), biphasic histologic features, and a prominent marginal zone (H&E, 100). B, White pulp with heterogeneous population of neoplastic cells or biphasic histologic features with central dark zone (arrowhead) and outer light zone (arrow) with transformed cells (H&E, 400). C, Red pulp, diffuse and nodular involvement (H&E, 400); inset, circulating villous lymphocyte (Wright-Giemsa, 1,650). D, Hilar lymph node showing preserved sinuses and prominent marginal zones (H&E, 50). deposits, present in the majority of cases, usually were in the central part of the nodules and varied from minimal deposition to coarse, broad bands. RP infiltration was nodular and diffuse in 9 of 13 cases Image 4C, only diffuse in 1 case, and only nodular in 3 cases. Epithelioid histiocytes, occasionally forming granulomas, were noted within the WP and RP in 3 cases (cases 34, 39, and 40), one of which had a history of multiple infections. Sections for a cyclin D1 stain were available for all cases except cases 33, 37, and 40. The EMZL with subsequent splenic involvement (case 35) showed morphologic features similar to those seen in the other SMZL cases in our series. Minor differences noted in this case included poorly circumscribed WP nodules with irregular, infiltrating borders and primarily diffuse RP infiltration with few RP nodules. The splenic HLN in this case also showed preservation of sinuses and a nodular infiltrate composed mainly of small lymphocytes with abundant cytoplasm (MZ morphologic features), similar to the lymph nodes from the SMZL cases. Residual or preserved mantle zones similar to those described in mucosa-associated lymphoid tissue lymphomas involving lymph nodes 18,19 were not noted on the H&E-stained sections of the HLN in this case. HLNs were available for review in all except 2 cases (cases 30 and 41). HLN sinuses were preserved Image 4D 342 Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB

9 Hematopathology / ORIGINAL ARTICLE in all except 2 cases (cases 39 and 42). The HLN in case 39 was small, with the architecture effaced by small lymphocytic cells in the absence of a distinct nodular pattern. MZs in HLNs (Image 4D) were present in 7 (78%) of 9 cases with MZs in the splenic sections. When present, MZs in HLNs typically were observed within the larger nodules. Rare residual GCs were identified in HLNs from 2 cases. Peripheral blood smears with circulating lymphoma cells were available for review in 7 cases (Image 4C, inset). The lymphoma cells usually were intermediate or large, with less condensed nuclear chromatin (compared with mature lymphocytes), single or multiple small nucleoli, and moderately abundant, light basophilic cytoplasm with or without villous cytoplasmic projections. Bone marrow morphology, available for review in 7 cases, usually showed a nodular and interstitial lymphoid infiltrate. Flow Cytometric Immunophenotyping The immunophenotypic features for the CLL/SLL and MZL cases were as described in the inclusion criteria. Paraffin section immunohistochemical stains were available in the MZL case without flow cytometric analysis (case 42). In MCL, the neoplastic cells showed a bright intensity of CD20 and surface immunoglobulin expression in all examined cases and were CD10, CD23, and FMC7+. The neoplastic cells in all cases except the blastoid variant of MCL (case 15) coexpressed CD5. The FL cells were CD5 CD20+ in all cases examined by flow cytometry, with CD10+ neoplastic cells in 5 of 5 cases. Immunophenotypic discordance in neoplastic cells was noted in 3 (33%) of 9 cases with more than 1 specimen analyzed by flow cytometry. In case 8, a subset of the peripheral blood lymphocytes was FMC7+ at the initial diagnosis of CLL, and splenectomy performed 6 years later showed the splenic lymphoma cells to be FMC7. The FL case without a previous LN biopsy (case 26) showed CD10 circulating lymphoma cells and CD10+ splenic lymphoma cells by flow cytometry. One SMZL case (case 37) showed dim CD10 expression in a small subset of the splenic cells by flow cytometry. However, the circulating lymphoma cells in this case were CD5, CD10, CD11c, CD23, CD25, and CD103 by flow cytometry, with morphologic features quite typical of SMZL and no circulating cleaved lymphocytes. Statistical Comparisons A summary of the results of statistical analyses is given in Table 3. In comparison with CLL/SLLs involving the spleen, MZLs involving the spleen are more likely to show MZs in the spleen, extracellular hyaline deposition in the splenic nodules, MZs in the HLNs, and preserved HLN sinuses. Residual GCs in the spleen seem to be more common in MZL than in CLL/SLL; however, this trend did not achieve statistical significance. MZLs in the spleen also are more likely than MCL to manifest with MZs in the HLNs, MZs in the spleen, residual GCs in neoplastic nodules in the spleen, and extracellular hyaline deposits in the spleen. However, preservation of HLN sinuses is unlikely to be helpful in differentiating MZL from MCL. In contrast, MZs in the spleen or in the splenic HLNs and extracellular hyaline deposits in the spleen are not helpful for discriminating MZL from low-grade FL involving the spleen. MZLs in the spleen are more likely than low-grade FLs to show residual GCs in the neoplastic nodules in the spleen and less likely than FLs to show effaced HLN sinuses. The presence of RP infiltration, either nodular or diffuse, is not helpful for discriminating MZL from any of the other studied SBL subtypes. Clinical Information The patients included 21 women (CLL/SLL, 3; MCL, 4; FL, 7; MZL, 7) and 21 men (CLL, 5; MCL, 5; FL, 5; MZL, 6), who underwent splenectomy at a median age of 65 years Table 3 Summary of Statistical Comparisons P * Values Variable MZL vs CLL/SLL MZL vs MCL MZL vs FL Splenic histologic features Residual germinal centers Extracellular hyaline Marginal zones Red pulp infiltration Nodular Diffuse Splenic hilar lymph nodes Marginal zones Preserved sinuses Sex CLL, B-cell chronic lymphocytic leukemia; FL, follicular lymphoma, grade 1 or 2; MCL, mantle cell lymphoma; MZL, marginal zone lymphoma. * P <.05 indicates a statistically significant difference between the lymphoma subtypes. Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB 343

10 Kansal et al / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS Table 4 Clinical Features of 42 Patients With Splenic Small B-Cell Lymphoma * Biopsy Sites Case No./Sex/ Outcome After Age (y) Diagnosis Stage BM LN Indication for Splenectomy Splenectomy 1/F/77 SLL IV + + Thrombocytopenia; 4 y after diagnosis AWD 2/M/69 CLL IV + + Thrombocytopenia; 5 y after diagnosis LFU 3/F/69 CLL III NA No biopsy AIHA; 2 y after diagnosis LFU 4/M/61 CLL IV + No biopsy Nondiagnostic; 1 y after BM biopsy DOD 5/F/82 CLL IV + + Thrombocytopenia DOD, <1 y 6/M/43 CLL/PL IV + No biopsy Diagnostic; thrombocytopenia, prominent splenomegaly, DOD, 3 mo 8 mo after BM biopsy 7/M/75 CLL IV + No biopsy Thrombocytopenia; 6 y after diagnosis DOD, 22 mo 8/M/64 CLL IV + No biopsy Pancytopenia, AIHA AWD, 4 wk 9/M/43 MCL IV + + Therapeutic, prominent splenomegaly AWD, 20 mo 10/F/68 MCL IV + + Thrombocytopenia AWD, 2 y 11/M/65 MCL IV + + Therapeutic; 4 mo after LN and BM biopsy diagnosis LFU, 2 y 12/F/59 MCL IV + + Therapeutic; within 6 mo of diagnosis NA 13/F/69 MCL IV + No biopsy, + Diagnostic; thrombocytopenia, prominent splenomegaly, DOD (by CT scan) 13 mo after BM biopsy 14/M/78 MCL IV + + Diagnostic; anemia, thrombocytopenia; 1 wk after BM DOD, 4 wk biopsy; LN biopsy 1 wk later 15/F/34 MCL, blastoid IV + No biopsy Therapeutic; tissue diagnosis 1 y after PB examination BMT, 1 mo; variant ANED, 39 mo 16/M/68 MCL IV + NA No biopsy Diagnostic; prominent splenomegaly 3 mo after PB Alive, 14 mo examination 17/M/73 MCL IV + + Diagnostic; before BM and LN biopsies Alive, 6 y 18/F/78 FL, grade 1 NA + Trauma; 2 y after BM biopsy AWD, 2 y 19/M/67 FL, grade 1 IV + + Relapse in spleen, thrombocytopenia, 14 mo after BM and CR, 4 y LN biopsies 20/F/62 FL, grade 1 IV + + Diagnostic; prominent splenomegaly AWD, 7 y 21/F/68 FL, grade 1 IV + + Diagnostic; 1 mo after BM biopsy AWD, 3 y 22/F/56 FL, grade 1 IV, liver + + Symptomatic relief Died, 18 mo 23/F/65 FL, grade 1 IV + + Nondiagnostic; 2 y after BM and LN biopsies Alive, 5 y 24/F/41 FL, grade 1 IV + + Diagnostic; concurrent BM and LN biopsies Alive, 5 y 25/M/53 FL, grade 1 IV + + Symptomatic relief; 3 y after LN biopsy Died, 16 mo 26/M/75 FL, grade 2 NA No biopsy Diagnostic; prominent splenomegaly, cytopenia Dead within 10 y 27/M/40 FL, grade 2 IV + + Symptomatic relief; 7 y after LN and BM biopsies Died, 1 y 28/M/55 FL, grade 2 IV + + Diagnostic; concurrent BM and abdominal LN biopsies Died, 6 mo 29/F/50 FL, grade 1 IV + + Diagnostic; 1 mo after BM biopsy; LN biopsy 3 y later Alive, 16 y 30/F/69 SMZL NA NA NA Diagnostic; AIHA NA 31/F/50 SMZL IV + + Diagnostic; cervical LN biopsy 10 y later, MZL Alive, 17 y 32/F/67 SMZL IV + No biopsy Diagnostic NA 33/F/62 SMZL IV + No biopsy Diagnostic; 2 mo after BM biopsy; pancytopenia Alive, 7 y 34/M/71 SMZL IV + Diagnostic; 6-mo history of pyrexia, chest infection Died, 1 y 35/F/56 EMZL IV + + Anemia; 14-y history of NHL in orbit and breast before AWD, 5 y splenectomy 36/M/41 SMZL IV + NA Diagnostic; 12-y history of splenomegaly and autoimmune Alive, 10 y symptoms after BMT 37/F/32 SMZL IV + No biopsy Diagnostic; 1 mo after PB examination Alive, 16 mo 38/M/67 SMZL IV + No biopsy Diagnostic; 3 y after BM biopsy Alive, 20 mo 39/M/51 SMZL IV + No biopsy Diagnostic; cytopenias; 1 y after BM biopsy Alive, 1 mo 40/M/56 SMZL IV + No biopsy Diagnostic; 3 mo after BM biopsy Alive, 9 mo 41/M/60 SMZL IV + No biopsy Diagnostic; 3 y after PB examination Alive, 1 mo 42/F/70 SMZL IV + Diagnostic; C1 esterase deficiency Alive, 11 mo AIHA, autoimmune hemolytic anemia; ANED, alive with no evidence of disease; AWD, alive with disease; BM, bone marrow; BMT, bone marrow transplant; CLL, B-cell chronic lymphocytic leukemia; CLL/PL, mixed CLL 16 or CLL with increased prolymphocytes 14 ; CR, complete remission; CT, computed tomography; DOD, died of disease; EMZL, extranodal MZL; FL, follicular lymphoma (World Health Organization classification); LFU, lost to follow-up; LN, lymph node; MCL, mantle cell lymphoma; MZL, marginal zone lymphoma; NA, not available; NHL, non-hodgkin lymphoma; PB, peripheral blood; SLL, small lymphocytic lymphoma; SMZL, splenic MZL; +, positive for lymphoma;, negative for lymphoma. * Age and stage at splenectomy. (range, years). The age ranges (median age) of the patients with each subtype of lymphoma were as follows: CLL/SLL, 43 to 82 years (69 years); MCL, 34 to 78 years (68 years); FL, 40 to 78 years (59 years); and MZL, 32 to 71 years (60 years). The individual clinical features, including stage of disease at splenectomy, presence of bone marrow involvement, lymph node involvement and biopsy, indications for splenectomy, and patient follow-up after splenectomy are given in Table 4. Splenectomy was performed primarily for diagnostic purposes in 23 patients (55%; SMZL, 12; FL, 6; MCL, 4; CLL with increased prolymphocytes, 1), who also had clinically prominent splenomegaly. 344 Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB

11 Hematopathology / ORIGINAL ARTICLE Splenectomy was performed primarily for therapeutic management (persistent cytopenia or reduction of tumor burden) in 18 patients (43%). Discussion Primary splenic lymphomas have been defined as lymphomas confined to the spleen and splenic HLNs. 4 In most instances, when an extensive search is made, the disease is found to be widespread, often involving the retroperitoneal lymph nodes or bone marrow. 6,7,20 According to a recent report of unselected and sequential splenectomies from 2 tertiary care centers, primary splenic lymphomas represent a clinical expression of a biologically heterogeneous group of lymphomas. 21 Our cases that underwent splenectomy for diagnostic subclassification do not represent strictly defined primary splenic lymphomas as defined by Kraus et al. 21 Instead, according to their criteria, these cases in our study may be considered as examples of permissively defined primary splenic lymphomas that manifested with prominent splenomegaly. Lymphomas manifesting with prominent splenomegaly may include SBLs such as SLL, FL, MCL, and MZL. 2,5,6,10,11,22,23 Differential diagnosis can be difficult because SBLs involving the spleen preferentially involve the WP 24,25 and may accentuate the normally present MZ structure in the spleen, thus mimicking a SMZL. Our study cases of SBL have well-characterized clinical, histologic, and immunophenotypic features and, thus, have a definitive diagnostic subclassification according to WHO criteria. 14 In general, we observed that the morphologic features of splenic WP nodules in well-characterized cases of SBL were a recapitulation of the histopathologic features of lymph nodes involved by the corresponding lymphoma. By gross examination, there is involvement of the WP, often in a miliary pattern and without the large, coalescent nodules that are characteristic of large cell lymphomas. MCL involving the spleen can lead to massive enlargement with larger macroscopic nodules than in other SBLs, correlating histologically with massive WP expansion and near-total obliteration of splenic architecture, which we noted in two thirds of our MCL cases. Diffuse RP infiltration was present in the leukemic phase of 2 MCL cases. Predominance of WP involvement, compared with that of RP involvement, was present in all cases. The splenic architecture was relatively preserved in all cases of SBL except in the cases of MCL with massive WP expansion. This architectural preservation may reflect the tendency of low-grade B-cell neoplasms to home to compartments of the spleen where their normal counterparts are found, in contrast with higher-grade B-cell lymphomas that do not respect the compartmental boundaries. 25 We observed residual GCs in WP nodules most frequently in SMZL cases and rarely in cases of MCL, FL, and CLL. Extracellular hyaline within WP nodules was common in both FL and SMZL. The MZ is a unique structure present around the lymphocytes of the mantle zone or primary follicles in the spleen and Peyer patches. 26,27 Lymphomas now presumed to arise from the MZ originally were described by Neiman et al. 28 The term splenic marginal zone lymphoma was coined later by Schmid et al. 29 Prominence of MZs may be seen in lymphomas other than MZL, 11,12,22 and we noted this frequently in our cases of FL. We also found that MCL can be a mimicker of the MZ pattern in the spleen, as previously described. 10,12 The differentiation of MCL from MZL often may be straightforward by evaluation of histopathologic and immunophenotypic features. 10,22,23,30 The characteristic histopathologic features of MCL with a monotonous small lymphocytic infiltrate contrast with those of MZL, which almost always has a polymorphous lymphoid population with scattered large lymphoid cells and is negative for cyclin D1 by immunohistochemical analysis. 30 In contrast, we found that distinguishing FL from SMZL can be difficult. One FL case in our series was reported previously as a pitfall in the diagnosis of SMZL. 31 At low magnification, FL involving the spleen can have a histologic appearance similar to that of MZL, with prominent WP expansion, extracellular hyaline deposits, often prominent MZs, and multiple small RP nodules similar to those described in MZL. 12 In addition, we observed MZs in HLNs, as well as in the spleen, in 40% (4/10) of our FL cases, which may add to the diagnostic dilemma. The critical diagnostic feature is the identification of neoplastic follicles within the WP or in HLNs, or even within the small RP nodules, in FLs involving the spleen. If necessary, applying the bcl-2 immunohistochemical stain to demonstrate neoplastic follicles in FL is helpful for differentiating them from reactive GCs in MZL. 22,23 Rarely, FLs involving the spleen also can have a few residual benign GCs, raising the differential diagnosis of reactive follicular hyperplasia. 32 We noted this pitfall in 1 case of FL in our study. The splenic morphologic features of the MZL cases in our study were similar to those described earlier. 29,33-35 The morphologic features of the circulating lymphoma cells were similar to those described earlier, with a subset of cases showing villous lymphocytes. 36 Clinically, we noted a history of an immunologic disorder in 25% of patients (3/12) with SMZL, unlike findings in a previous study in which no such history was found in 12 patients with SMZL 37 and similar to findings in a later report. 38 The morphologic features of the neoplastic nodules in lymph nodes involved by SMZL have been reported to resemble those of the WP in that entity. 18 Our findings in HLNs from 34 splenic SBL Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB 345

12 Kansal et al / HISTOPATHOLOGIC FEATURES OF SPLENIC SMALL B-CELL LYMPHOMAS cases confirm the similarity in morphologic features in HLNs and splenic nodules in all SBL subtypes. Furthermore, we found that unequivocal neoplastic follicles in HLNs with effaced nodal sinuses is a most helpful feature in the diagnosis of FL and for distinguishing FL from MZL, which almost always had preserved nodal sinuses. In the study by Mollejo et al, 18 HLNs from all cases of SMZL had preservation of sinuses, similar to the findings in HLNs from most of our MZL cases. We observed a MZ pattern on H&E-stained sections in splenic HLNs from 80% (8/10) of our SMZL cases, compared with 1 of 14 HLN cases reported by Mollejo et al. 18 We did not detect a MZ pattern in our CLL/SLL cases. Evaluation of a splenectomy specimen in these cases usually is not difficult, since the diagnosis usually is established before splenectomy, which is most likely performed for improvement of cytopenias. Furthermore, the cytologic features of the larger WP nodules in spleens involved by CLL/SLL may be typical of lymph nodes involved by CLL/SLL, with the presence of proliferation centers or scattered paraimmunoblasts. From our study, we conclude that well-characterized cases of SBLs involving the spleen often have typical morphologic features, with the morphologic features of the splenic WP and RP nodules reminiscent of the histopathologic features of lymph nodes involved by the corresponding lymphoma. Histopathologic evaluation of HLNs is invaluable in differential diagnosis, since obliteration or effacement of architecture caused by SBL may be easier to interpret in lymph nodes than in the spleen. Therefore, a search for HLNs must always be performed in splenectomy specimens. MZs may be observed in the spleen and in splenic HLNs involved by SBLs other than SMZL. Nevertheless, precise classification of SBLs in a splenectomy specimen can be achieved by careful assessment of histopathologic features in most cases. In comparison with FL, MZLs in the spleen are more likely to show residual GCs in the neoplastic WP nodules in the spleen (P =.000) and less likely to show effaced HLN sinuses (P =.000), while MZs in the spleen or in the HLNs are not significantly different in MZL and FL. MZL and MCL can be distinguished by the cytologic composition of the neoplastic infiltrate and cyclin D1 immunohistochemical analysis. Furthermore, in comparison with MCL, MZLs are more likely to show MZs in splenic HLNs (P =.002), MZs in the spleen (P =.026), residual GCs in the spleen (P =.026), and extracellular hyaline deposits (P =.044), while preservation of HLN sinuses is not significantly different in MZL and MCL. In comparison with CLL/SLL, MZLs are more likely to show MZs in the spleen (P =.000), MZs in HLNs (P =.002), and preservation of HLN sinuses (P =.002). RP infiltration, nodular or diffuse, is not significantly different in MZL compared with CLL/SLL, MCL, or low-grade FL. In difficult cases, overall assessment and correlation of the histologic features with peripheral blood and bone marrow morphologic, immunophenotypic, and clinical features is most helpful and usually leads to precise subclassification of SBLs involving the spleen. From the Department of Pathology, University of Michigan, Ann Arbor. Address correspondence to Dr Kansal: Dept of Pathology, 7th Floor, Buffalo General Hospital, 100 High St, Buffalo, NY References 1. Kim H, Dorfman RF. Morphological studies of 84 untreated patients subjected to laparotomy for the staging of non- Hodgkin s lymphomas. Cancer. 1974;33: Warnke RA, Weiss LM, Chan JKC, et al. Lymphomas of the spleen. In: Tumors of the Lymph Nodes and Spleen. Washington, DC: Armed Forces Institute of Pathology; 1995: Rosai J, ed. Atlas of Tumor Pathology; Third series, Fascicle Ahmann DL, Kiely JM, Harrison EG, et al. Malignant lymphoma of the spleen: a review of 49 cases in which the diagnosis was made at splenectomy. Cancer. 1966;19: Brox A, Bishinsky JI, Berry G. Primary non-hodgkin lymphoma of the spleen. Am J Hematol. 1991;38: Kraemer BB, Osborne BM, Butler JJ. Primary splenic presentation of malignant lymphomas and related disorders: a study of 49 cases. Cancer. 1984;54: Narang S, Wolf BC, Neiman RS. Malignant lymphoma presenting with prominent splenomegaly: a clinicopathologic study with special reference to intermediate cell lymphoma. Cancer. 1985;55: Spier CM, Kjeldsberg CR, Eyre HJ, et al. Malignant lymphoma with primary presentation in the spleen: a study of 20 patients. Arch Pathol Lab Med. 1985;109: Jaffe ES, Raffeld M, Medeiros LJ. Histopathologic subtypes of indolent lymphomas: caricatures of the mature B-cell system. Semin Oncol. 1993;20: Mason DY, Harris NL, eds. Human Lymphoma: Clinical Implications of the REAL Classification. London, England: Springer-Verlag; 1999: Pittaluga S, Verhoef G, Criel A, et al. Small B-cell non- Hodgkin s lymphomas with splenomegaly at presentation are either mantle cell lymphoma or marginal zone cell lymphoma: a study based on histology, cytology, immunohistochemistry, and cytogenetic analysis. Am J Surg Pathol. 1996;20: Arber DA, Rappaport H, Weiss LM. Non-Hodgkin s lymphoproliferative disorders involving the spleen. Mod Pathol. 1997;10: Piris MA, Mollejo M, Campo E, et al. A marginal zone pattern may be found in different varieties of non-hodgkin s lymphoma: the morphology and immunohistology of splenic involvement by B-cell lymphomas simulating splenic marginal zone lymphoma. Histopathology. 1998;33: Harris NL, Jaffe ES, Stein H, et al. A revised European- American classification of lymphoid neoplasms: a proposal from the International Lymphoma Study Group. Blood. 1994;84: Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB

13 Hematopathology / ORIGINAL ARTICLE 14. Jaffe ES, Harris NL, Stein H, et al, eds. Pathology and Genetics of Tumours of Hematopoietic and Lymphoid Tissues. Lyon, France: IARC Press; 2001: World Health Organization Classification of Tumours. 15. Tworek JA, Singleton TP, Schnitzer B, et al. Flow cytometric and immunohistochemical analysis of small lymphocytic lymphoma, mantle cell lymphoma, and plasmacytoid small lymphocytic lymphoma. Am J Clin Pathol. 1998;110: Bennett JM, Catovsky D, Daniel M-T, et al. Proposals for the classification of chronic (mature) B and T lymphoid leukemias. J Clin Pathol. 1989;42: Melo JV, Catovsky D, Galton DA. The relationship between chronic lymphocytic leukaemia and prolymphocytic leukaemia. Br J Haematol. 1986;63: Mollejo M, Lloret E, Menarguez J, et al. Lymph node involvement by splenic marginal zone lymphoma: morphological and immunohistochemical features. Am J Surg Pathol. 1997;21: Campo E, Miquel R, Krenacs L, et al. Primary nodal marginal zone lymphomas of splenic and MALT type. Am J Surg Pathol. 1999;23: Falk S, Stutte HJ. Primary malignant lymphomas of the spleen. Cancer. 1990;66: Kraus MD, Fleming MD, Vonderheide RH. The spleen as a diagnostic specimen: a review of 10 years experience at two tertiary care institutions. Cancer. 2001;91: Delsol G, Diebold J, Isaacson PG, et al. Pathology of the spleen: report on the workshop of the VIIIth meeting of the European Association for Haematopathology, Paris Histopathology. 1998;32: Isaacson PG. Malignant lymphomas with follicular growth pattern. Histopathology. 1996;28: Van Krieken JHJM, Feller AC, te Velde J. The distribution of non-hodgkin s lymphoma in the lymphoid compartments of the human spleen. Am J Surg Pathol. 1989;13: Van Krieken JHJM. Histopathology of the spleen in non- Hodgkin s lymphoma. Histol Histopathol. 1990;5: Van Krieken JHJM, te Velde J. Normal histology of the human spleen. Am J Surg Pathol. 1988;12: Van Krieken JHJM, von Schilling C, Kluin M, et al. Splenic marginal zone lymphocytes and related cells in the lymph node: a morphologic and immunohistochemical study. Hum Pathol. 1989;20: Neiman RS, Sullivan AL, Jaffe R. Malignant lymphoma simulating leukemic reticuloendotheliosis: a clinicopathologic study of ten cases. Cancer. 1979;43: Schmid C, Kirkham N, Diss T, et al. Splenic marginal zone cell lymphoma. Am J Surg Pathol. 1992;16: Savilo E, Campo E, Mollejo M, et al. Absence of cyclin D1 protein expression in splenic marginal zone lymphoma. Mod Pathol. 1998;11: Alkan S, Ross CW, Hanson CA, et al. Follicular lymphoma with involvement of the splenic marginal zone: a pitfall in the differential diagnosis of splenic marginal zone cell lymphoma. Hum Pathol. 1996;27: Burke J. Surgical pathology of the spleen: an approach to the differential diagnosis of splenic lymphomas and leukemias. Am J Surg Pathol. 1981;5: Hammer RD, Glick AD, Greer JP, et al. Splenic marginal zone lymphoma. Am J Surg Pathol. 1996;20: Isaacson PG, Matutes E, Burke M, et al. The histopathology of splenic lymphoma with villous lymphocytes. Blood. 1994;84: Mollejo M, Menarguez J, Lloret E, et al. Splenic marginal zone lymphoma: a distinctive type of low-grade B-cell lymphoma. Am J Surg Pathol. 1995;19: Catovsky D, Matutes E. Splenic lymphoma with circulating villous lymphocytes/splenic marginal zone lymphoma. Semin Hematol. 1999;36: Dierlamm J, Pittaluga S, Wlodarska I, et al. Marginal zone B- cell lymphomas of different sites share similar cytogenetic and morphologic features. Blood. 1996;87: Murakami H, Irisawa H, Saitoh T, et al. Immunologic abnormalities in splenic marginal zone cell lymphoma. Am J Hematol. 1997;56: Am J Clin Pathol 2003;120: DOI: /HWG084N3F3LRJ8XB 347